JPWO2006013915A1 - Display panel inspection method, inspection apparatus, and manufacturing method - Google Patents

Abstract

The height of the substrate surface including the liquid material application part while having the height measurement means and moving the substrate or the height measurement means in a direction intersecting the liquid material applied to the substrate at a predetermined interval. Perform the measurement discretely, extract the height of each liquid material from the height shape signal obtained by approximating between the obtained discrete height shape signals, and use the height signal as a test signal. A display panel inspection method, an inspection apparatus, and a manufacturing method using them, characterized by measuring a coating amount for each liquid material. Immediately after the phosphor paste application process, the condition of the application process is inspected to quickly find defects that cause continuous defects in the application process, minimizing the number of substrates that become defective and lost, and promptly It makes it possible to restore the process.

Description

  The present invention relates to a display panel inspection method, an inspection apparatus, and a manufacturing method, and more particularly, in a display panel in which a plurality of liquid materials are applied on a substrate, means used for accurately applying and forming the liquid material on the substrate. The present invention relates to a state inspection method and an inspection apparatus, and a display panel manufacturing method using the method and apparatus.

  In order to construct a phosphor on a substrate of a display panel, a technique for inspecting the configuration state of the phosphor layer in a method of applying a plurality of liquid materials on the substrate is known. For example, Patent Document 1 discloses a technique for inspecting the configuration state of a phosphor layer from the difference between the substrate surface shape before the phosphor composition and the substrate surface shape on which the phosphor layer is constructed after the firing process after applying the phosphor paste. Is disclosed. However, in this method, since the measurement is performed twice for one product, the manufacturing cost is increased. In addition, since the quality of the product cannot be judged until the second measurement is performed, if a continuous failure occurs in the phosphor paste coating device (and firing furnace), a large number of defective substrates are generated. Will do.

In Patent Document 1, ultraviolet rays are applied to the substrate surface on which the phosphor layer is formed through a baking process after applying the phosphor paste, and each phosphor of R (red), G (green), and B (blue) is used. A technique for inspecting the configuration state of the phosphor layer by measuring the amount of excited luminescence from the layer is also disclosed. However, in this method, since the quality of the product cannot be determined until the measurement after firing is performed, if a defect that causes a continuous defect occurs in the phosphor paste coating apparatus, a large number of substrates that are also defective products are also produced. Will occur.
Japanese Patent Laid-Open No. 9-273913

  It is an object of the present invention to quickly find defects that cause continuous defects in the coating process by inspecting the state of the coating process immediately after the phosphor paste coating process, and minimize the number of substrates that become defective products and loss. It is an object of the present invention to provide a display panel inspection method and inspection apparatus, and a manufacturing method using them, which can be reduced to the limit and promptly restore the process. Another object is to make it possible to obtain a desired measurement result in a single measurement and to reduce the manufacturing cost of the substrate. Furthermore, another object is to manage substrate surface state data so that it can be used as data for manufacturing products with higher accuracy and higher quality.

  In order to solve the above-described problems, a display panel inspection method according to the present invention includes a height measuring unit, and a substrate or a height in a direction intersecting with a plurality of liquid materials applied to the substrate at a predetermined interval. While moving the height measuring means, discretely measure the height of the substrate surface including the liquid material application part, and obtain the approximate curve from the obtained discrete height shape signal, and then obtain the liquid material from the height shape signal obtained The height signal obtained by extracting each height is used as an inspection signal, and the coating amount for each liquid material is measured from the inspection signal.

  In this inspection method, the signal of the liquid phosphor application part is specified from the discrete height shape signal obtained by the height measuring means, and the height shape signal is obtained from the specified signal using a conic curve as an approximate curve. Can be.

  In this inspection method, the signal of the liquid phosphor coating part is specified from the discrete height shape signal obtained by the height measuring means, and the height shape signal is obtained from the specified signal using a circle as an approximate curve. The approximate circle diameter signal obtained by connecting the diameters of the approximate circles so as to correspond to a plurality of liquid materials can be used as an inspection signal, and the application amount for each liquid material can be measured from the inspection signal.

  In the inspection method, a plurality of first barrier ribs are formed on the substrate at predetermined intervals in a direction parallel to the longitudinal direction of the liquid material applied at predetermined intervals, and further adjacent to each other. The present invention can be suitably applied to a substrate in which a plurality of second partition walls are formed at a predetermined interval in a direction perpendicular to the longitudinal direction of the liquid material between one partition wall.

  With respect to the substrate having such a configuration, a height measuring sensor having a spot-shaped measuring region is used as a height measuring means, and a central portion between the second partition walls formed in the direction perpendicular to the longitudinal direction of the liquid material ± The shape of the region within 35% can be measured over the entire length of the substrate in a direction transverse to the longitudinal direction of the liquid material.

  In this method, there is further provided a substrate position restricting means for restricting the position of the substrate, and the shape of the region within ± 35% of the central portion between the second partition walls is formed over the entire length of the substrate in the direction crossing the longitudinal direction of the liquid material. Can be measured. In other words, the position of the substrate is regulated (conveyance guide in the case of substrate movement, pre-positioning mechanism in the case of sensor movement), and height measurement sensor scanning in the region within ± 35% of the center between the second partition walls is realized. It is.

  The method further includes a substrate position recognizing unit for recognizing the position of the substrate and a scanning position correcting unit for correcting the position of the height measuring unit based on the substrate position information. The shape of the region within ± 35% of the portion can be measured over the entire length of the substrate in a direction crossing the longitudinal direction of the liquid material. In other words, by measuring the substrate edge position, the substrate tilt / meander information is obtained, the height measurement sensor position is corrected, and the height measurement sensor scanning within the region within ± 35% of the center between the second partition walls is realized. To do.

  Further, the method further includes two or more height measuring means, a position adjusting means, and a switching means, and the shape of the region within ± 35% of the central portion between the second partition walls is defined by the length of the liquid material. It is possible to measure over the entire length of the substrate in a direction crossing the direction. That is, at least two or more height measurement sensors are used, and at least one height measurement sensor obtains data in a region within ± 35% of the central portion between the second partition walls even if the substrate tilts or meanders. Is.

  Alternatively, a height measurement sensor having a measurement region including a second partition wall interval formed in a direction perpendicular to the longitudinal direction of the liquid material is used as the height measurement means, and the shape of the substrate surface is crossed across the longitudinal direction of the liquid material. It is also possible to measure over the entire length of the substrate in the direction.

  In the display panel inspection method according to the present invention as described above, it has a substrate back surface height measuring means for measuring the height of the substrate back surface, and the measurement result by the height measuring means is corrected with the substrate back surface height measurement result. You can also do it. That is, it has a second height measuring sensor for measuring the back surface of the substrate, measures the vertical movement of the substrate, and eliminates the influence of the vertical movement of the substrate from the height measurement data.

  Further, the measurement position of the height measuring means can be arranged at a position where the substrate moving means and the substrate are in contact with each other. That is, the measurement position of the height measuring means is arranged at a position where the substrate carrying means and the substrate are in contact with each other, thereby suppressing the vertical movement of the substrate.

  Further, in the above inspection method, the liquid material applied at a predetermined interval changes its surface shape between the first and second partition walls by a fluid action immediately after application, and reaches a steady state after a predetermined time. In this case, the height of the substrate surface can be measured after a predetermined time. That is, the inspection is performed after the liquid material is leveled.

  In addition, when the liquid material applied at a predetermined interval changes the surface shape between the first and second partition walls by a fluid action immediately after the application and reaches a steady state after a predetermined time, the liquid material with respect to time It is also possible to correct the height shape signal with the surface shape change information. That is, the height shape signal is corrected using paste leveling characteristic data with respect to the time measured in advance.

  Further, in the inspection method, in the signal processing step of providing a predetermined defect determination threshold for determining the presence or absence of defects in the inspection signal, each of the regions corresponding to a plurality of liquid materials to be measured in the inspection signal is specified, A specific defect determination threshold value can be provided for each identified signal portion.

  In this case, the defect determination threshold for the inspection signal can be automatically adjusted from the moving average signal obtained by performing the moving average process on the inspection signal itself obtained from the inspection target substrate.

  In addition, the height of the substrate surface is continuously measured for a plurality of substrates, and the defect determination threshold of the inspection target substrate is determined from the height shape information of the substrate measured before the measurement of the inspection target substrate. It can also be adjusted automatically. That is, an individual determination threshold is automatically set for each liquid material from the measurement result of another substrate performed before the measurement of the target substrate.

  In the inspection method, the height measurement is performed on all the substrates to which the liquid material is applied every time the liquid material is applied to the substrate, or after the liquid material is applied to a plurality of substrates. It can be performed on all substrates coated with liquid material or on selected representative substrates. For example, the inspection timing and the target substrate are selected from the accuracy of inspection, manufacturing tact, the number of lost substrates when NG (no good) occurs, etc.

  In the above inspection method, height measurement information obtained from a plurality of substrates can be managed and fed back to the control and operation of the coating apparatus.

  The display panel inspection apparatus according to the present invention has a height measuring means for discretely measuring the height of the substrate surface including the liquid material application portion, and an approximate curve is obtained from the obtained discrete height shape signal to obtain a height. It comprises signal processing means for obtaining a shape signal.

  In this inspection apparatus, a moving means for moving the substrate or the height measuring means in a direction intersecting with a plurality of liquid materials applied to the substrate at predetermined intervals, and a measurement result and an inspection result by the signal processing means are output. The information output means can be further provided.

  Further, the substrate fixing means for fixing the substrate may be provided, and the substrate fixing means may have a position correcting function in the rotation direction with the axis perpendicular to the substrate surface as the central axis.

  Also, a laser displacement meter is used as the height measuring means, a linear motor guide with an air bearing is used as the moving means for moving the height measuring means, and a vertical axis is centered on the substrate surface as the substrate fixing means for fixing the substrate. It can be set as the form comprised using the high precision stage which has a position correction function of a rotation direction as an axis | shaft.

  In this case, a high-precision stage as a substrate fixing means for fixing the substrate can be used in common with the coating apparatus as the substrate fixing means when the liquid material is applied. A general-purpose stage can be used as the high-precision stage.

  The inspection apparatus may further include a substrate position restricting unit for restricting the position of the substrate.

  In this case, a laser displacement meter can be used as the height measuring means, a roller transport machine can be used as the moving means for moving the substrate, and a position regulation guide can be used as the substrate position regulating means.

  Further, a laser displacement meter can be used as the height measuring means, a single axis stage can be used as the moving means for moving the height measuring means, and a positioning mechanism can be used as the substrate position regulating means.

  The inspection apparatus may further include a position correction unit for correcting the positions of the substrate edge position measurement unit and the height measurement unit.

  In this case, a laser displacement meter is used as the height measuring means, a roller transport machine is used as the moving means for moving the substrate, a laser position measuring sensor is used as the substrate edge position measuring means, and a single axis stage is used as the position correcting means. Can be configured.

  The inspection apparatus may further include an installation interval adjusting unit that adjusts an installation interval between at least two height measuring units and the height measuring units.

  In this case, two laser displacement meters can be used as the height measuring means, a roller transport machine can be used as the moving means for moving the substrate, and a single-axis stage can be used as the installation interval adjusting means.

  Such an inspection apparatus further includes a substrate back surface height measuring unit, and a laser displacement meter can be used as the substrate back surface height measuring unit.

  Further, the laser displacement meter as the height measuring means can be configured to measure the position where the substrate moving means and the substrate are in contact with each other.

  The display panel manufacturing method according to the present invention includes a method characterized by manufacturing a display panel using the above inspection method or the above inspection apparatus.

  In this manufacturing method, the substrate can be corrected using the liquid material correcting means based on the defect information of the substrate. That is, the NG substrate is corrected based on the defect information of the substrate.

  According to the present invention, since a defect in the coating process (clogging of the coating nozzle, etc.) can be detected immediately after the occurrence of the defect from the surface shape of the substrate, a substrate that is lost due to the occurrence of a defective product (hereinafter referred to as an NG loss substrate). The number of sheets can also be minimized. Further, since the measurement only needs to be performed once by the displacement meter, an increase in manufacturing cost can be minimized.

  Even if the interval of the discrete height shape signals is increased in order to shorten the time required for the inspection, the height shape signals can be obtained with high accuracy by approximation, so that high measurement accuracy can be maintained. Further, even when low frequency noise is generated in the measured discrete height shape signal, the noise can be removed without losing the shape information on the surface of the liquid material, and high measurement accuracy can be maintained.

  High inspection sensitivity can be obtained by estimating the filling amount by the radius of the approximate circle including the surface of the liquid material (especially in the manufacturing specification in which the liquid material is filled in the partition wall).

  Even if it is a substrate in which a plurality of second partition walls are formed at predetermined intervals in the direction perpendicular to the longitudinal direction of the liquid material between adjacent first partition walls, a so-called substrate with lateral ribs, Highly accurate inspection is possible.

  By specifying the accuracy of scanning, the accuracy / reliability of inspection can be improved.

  Scanning accuracy can be ensured by a travel guide in the case of substrate movement and a substrate positioning mechanism in the case of sensor movement.

  By making the measurement visual field follow the sensor, the scanning accuracy can be secured.

  Scanning accuracy can be ensured by one of the two sensors always measuring the measurement visual field.

  By performing the height measurement within the wide field of view and performing the inspection with the average height within the field of view, it is possible to perform a high-precision inspection even with the ancestor precision scanning.

  By correcting the height signal with the substrate back surface height measurement signal, the substrate vertical movement noise can be eliminated.

  By setting the measurement field of view on the roller, the vertical movement of the substrate can be suppressed, and the vertical movement noise of the substrate can be eliminated.

  The accuracy of the inspection can be improved by performing the inspection after waiting for the leveling of the liquid material (paste).

  By correcting the measurement data with the leveling characteristic of the paste with respect to time, the accuracy of the inspection can be increased.

  It is also possible to perform inspection by manually eliminating individual differences of coating apparatuses and fixed manufacturing unevenness of the substrate.

  Individual inspection of the coating apparatus and fixed manufacturing unevenness of the substrate can be automatically eliminated by a spatial moving average process to perform inspection.

  It is also possible to perform inspection by automatically eliminating individual differences of coating apparatuses and fixed manufacturing unevenness of the substrate by temporal moving average processing.

  For a multi-planar substrate, the inspection timing and the target substrate can be selected from the number of loss substrates when NG occurs, the manufacturing tact, the accuracy of inspection, and the like.

  The surface shape data of the substrate measured for the inspection of the state of the coating process can be managed as a trend and fed back to the control and operation of the coating process to enable stable substrate production.

  Actually, the inspection apparatus can be constituted by the height measuring means and the signal processing means.

  Further, a specific inspection apparatus can be actually configured by including a moving means for moving the height measuring means and an output means for outputting the inspection result.

  Furthermore, a specific inspection apparatus can be actually configured by including a substrate fixing means having a function of correcting the rotation direction (θ direction) of the substrate.

  In this way, a more specific device configuration is possible.

  In addition, by incorporating this inspection apparatus in the coating machine, the occurrence of defects can be found more quickly.

  By providing the above apparatus configuration with a substrate position regulating means, a more specific inspection apparatus can be configured in practice.

  In the configuration provided with the substrate position regulating means, a substrate moving type device configuration is possible, and a height measuring means moving type device configuration is also possible.

  In addition, the above-described apparatus configuration can further include a substrate edge position measuring unit and a height measuring unit to actually configure a specific inspection apparatus.

  In the configuration including the substrate edge position measuring means and the height measuring means, a substrate moving type apparatus configuration is possible.

  In addition, by using two height measuring means in the above apparatus configuration and further including an interval adjusting means for adjusting the distance between the height measuring means, a specific inspection apparatus can be configured in practice.

  Also in this inspection apparatus, a substrate movement type apparatus configuration is possible.

  Furthermore, a specific inspection apparatus can be actually configured by providing the above-described apparatus configuration with a substrate back surface height measuring means.

  Further, in the inspection apparatus as described above, a configuration capable of suppressing the vertical movement of the substrate is possible.

  By incorporating this inspection apparatus in the liquid material coating apparatus, the occurrence of defects can be detected more quickly.

  Since it is possible to detect a coating process defect (coating nozzle clogging, etc.) immediately after the occurrence of a defect from the surface shape of the substrate, it is possible to realize a substrate manufacturing method and apparatus capable of minimizing the number of NG loss substrates. Since the measurement only needs to be performed once by the displacement meter, an increase in the manufacturing cost of the display panel can be minimized.

  It is possible to increase the yield of the entire process by correcting the substrate at the time of NG in the correction process with respect to the substrate measured for the state inspection of the coating process.

It is a schematic block diagram which shows the structure of PDP. It is a process flowchart which shows a PDP backplate manufacturing process. It is a schematic partial perspective view which shows the PDP backplate with which the fluorescent substance is unconfigured. It is a schematic partial perspective view which shows the example of the PDP backplate immediately after fluorescent substance paste application | coating. It is a schematic partial perspective view which shows the example of the PDP backplate after the fluorescent substance paste is leveled. It is a general | schematic fragmentary perspective view which shows the example of the PDP backplate with which the fluorescent substance layer was comprised (only 1 color). It is a schematic perspective view which shows the relationship between a PDP backplate and height measurement means scanning. It is the schematic which shows the relationship of the sampling of a PDP backplate surface shape and a height measurement means. It is explanatory drawing explaining the definition of a signal approximation method and height h / approximate circle radius r. It is explanatory drawing explaining a height shape signal / height signal / approximate circle radius signal, and a defect determination threshold value. It is explanatory drawing explaining the sensitivity characteristic of a height signal test | inspection and an approximate circle signal test | inspection. It is the schematic which shows a PDP backplate application direction surface shape and a spot measurement position. It is the schematic which shows the PDP backplate application direction surface shape and the width measurement position. It is explanatory drawing explaining the relationship between a spot / wide measurement position and inspection sensitivity. It is explanatory drawing explaining the leveling phenomenon of a fluorescent substance paste. It is explanatory drawing explaining the relationship between the elapsed time after fluorescent substance paste application | coating, and surface height. It is explanatory drawing explaining the fixed threshold value and individual threshold value in a test | inspection signal. It is explanatory drawing explaining the fixed threshold value and automatic fluctuation | variation threshold value in a test | inspection signal. It is explanatory drawing explaining the difference threshold value in the difference process waveform of a test | inspection signal. It is a schematic block diagram which shows the inspection apparatus integrated in the same body as the coating device. It is a schematic plan view which shows the board | substrate conveyance and board | substrate stop by a roller conveyance machine. It is a schematic plan view of the roller transporter which shows the board | substrate movement type measuring apparatus provided with the board | substrate position control means. It is a schematic plan view of the roller transporter which shows the sensor movement type measuring apparatus provided with the board | substrate position control means. It is a schematic plan view of the roller transport machine which shows the inspection apparatus provided with the board | substrate position recognition means and the position correction means. It is a schematic plan view of the roller transport machine which shows the inspection apparatus provided with two height measurement means and a space | interval adjustment means. It is a schematic block diagram which shows the inspection apparatus provided with the board | substrate back surface height measurement means. It is a schematic block diagram which shows the inspection apparatus in which the measurement point of the height measurement means was installed in the contact point of a board | substrate and a board | substrate conveyance means. It is a schematic block diagram which shows the inspection apparatus provided with two height measurement means and a space | interval adjustment means. It is explanatory drawing explaining the measurement error of the height signal after correction | amendment by a discrete height signal and a conical curve. It is explanatory drawing explaining the measurement result at the time of with / without correction | amendment by a conical curve, and a measurement error. It is explanatory drawing explaining the measurement error of the height signal value after correction | amendment by the height signal value after correction | amendment by a discrete height signal and a moving average process, and a conical curve. It is explanatory drawing explaining the measurement result at the time of correction | amendment execution by a moving average process, and the time of correction | amendment implementation by a conical curve.

Explanation of symbols

1 Back plate 2 Front plate 1a Mother glass substrate 1b before application process Mother glass substrate 1b1 in application process PDP back plate 1 in mother glass
1b2 PDP back plate 2 in the mother glass
1b3 PDP back plate 3 in mother glass
1b4 PDP back plate 4 in the mother glass
1b5 PDP back plate 5 in the mother glass
1b6 PDP back plate 6 in the mother glass
1c Mother glass substrate 1d after coating process PDP back plate 11 after coating process Bulkhead (vertical rib)
12 Address electrode 12a Electrode a
13 Back glass substrate 14 Dielectric layer 15 Discharge space 16 Bulkhead (lateral rib)
17 Grooves with lateral ribs 17r R grooves with lateral ribs 17g G grooves with lateral ribs 17b B grooves with lateral ribs 18 Cell 19 Coating direction 21 Front glass substrate 22 Dielectric layer 23 Display electrode 24 Protective film 31 Cleaning / drying process 32 Pattern Electrode formation process 33 Dielectric layer formation process 34 Partition formation process 35 Phosphor coating process 36 Coating process state inspection process 37 Phosphor drying process 38 Defect correction process 40b B phosphor paste (normal)
40b 'B phosphor paste (abnormal)
40b "B coating missing 41b B phosphor after drying (normal) (after leveling)
41b 'B phosphor after drying (abnormal) (after leveling)
41b "B coating missing after drying (after leveling)
42r R phosphor 42g G phosphor 42b B phosphor after drying (normal)
42b 'B phosphor after drying (abnormal)
42b "B coating missing after drying 43 Paste flow 50 Height measuring means 50a Spot displacement sensor 50a 'Second spot displacement sensor 50b Wide displacement sensor 51 Measuring light 51a Scanning trajectory 51a' of spot displacement sensor Second spot displacement Sensor scanning trajectory 52 Sampling timing 60a Discrete height shape signal 60b outside approximate region Discrete height shape signal 60c within approximate region Approximate circle 60d Height shape signal within approximate region (approximated)
61 Discrete height shape signal (60a + 60b)
62 Height shape signal (60a + 60d)
63 Height signal 63b Height of B phosphor (normal) 63b ′ Height of B phosphor (abnormal) 63b ″ Height of B phosphor coating missing portion 64 Approximate circular diameter signal 64b Approximation of B phosphor (normal) Circular radius 64b 'Approximate circular radius 64b "of B phosphor (abnormal) B Approximate circular radius 70 of phosphor coating drop-out portion 70 Substrate fixing means 71 Moving means 72 Height measuring means moving means 73 Coating means fixing means 74 Coating means 75L Substrate Loading means 75UL Substrate unloading means 76 Height measuring means fixing means 77 Inspection apparatus operation section 78 Coating apparatus operation section 100 PDP
101 Plasma 102 Display light 200 Axis 201 Roller 202 Roller shaft 203 Substrate transport direction 203 'Sensor movement direction 203 "Correction direction 203"' Spacing adjustment direction 204 Shaft perpendicular to substrate surface 205 Rotation direction tilt 206 Substrate transport means 220 Substrate transport Substrate position restricting means 230 Substrate position restricting means 240 while the substrate is stopped Substrate position recognizing means 241 Scanning position correcting means 250 Interval adjusting means 251a Measurement area a
251b Measurement area b
251c Measurement area c
260a Height measuring means moving means 260b Back surface height measuring means moving means 261 Space 262 Contact area 280 of the substrate moving means and the substrate 280 Fixing means 281 Inspection device operation unit 282 Cable c1 Horizontal rib direction cross section base point c1 'Horizontal rib direction cross section Line end point c2 Vertical rib direction cross section line (normal application) base point c2 ′ Vertical rib direction cross section line (normal application) end point c3 Vertical rib direction cross section line (abnormal application) base point c3 ′ Vertical rib direction cross section line (abnormal application) end point d1 Scan d2 Oblique scan dw Approximate area Em Measurement error h Paste height (PL-KL)
KL Reference surface level PL Paste surface level PL0 Paste surface level PL0 'at the cell center (P0) Paste surface level PL2 at the cell center (P0) after leveling Paste surface level PL2' on the lateral rib (P2) after leveling Paste surface level p0 on horizontal rib (P2) Sensor visual field center position (cell center)
p1 Sensor visual field center position (cell edge)
p2 Sensor visual field center position (on horizontal rib)
r Approximate circle radius s M + 3rd paste height (OK)
sw1 sensor scanning width sw1 'sensor scanning width sw1 "sensor scanning width sw2 sensor field width t M + 6th paste height (OK)
t ”M + 6th paste height (NG)
thr Threshold value thr1 in height signal First (lower) threshold value thr2 in approximate circle radius Second (upper) threshold value u M + 9th paste height in approximate circle radius (OK)
u ”M + 9th paste height (NG)
v M + 3rd paste height of Nth substrate (OK)
v 'N + 1 M + 3rd paste height of the 1st substrate (OK)
v ”M + 3rd paste height of N + 2nd substrate (OK)
w M + 6th paste height of Nth substrate (OK)
W 'N + 1 M + 6th paste height on the 1st substrate (OK)
w ”N + 6th paste height of N + 2nd substrate (NG)
x M + 9th paste height of Nth substrate (OK)
x 'N + 1 M + 9th paste height on the 1st substrate (OK)
x ”M + 9th paste height of N + 2nd substrate (NG)
Y Nth measurement signal Y ′ in the discrete measurement signal waveform N ′ N + 1th measurement signal in the discrete measurement signal waveform (a) Liquid material lowest part measurement result with conic curve approximation when the liquid material filling condition is changed (b) Liquid material lowest part measurement result without conic curve approximation when changing liquid material filling condition (c) Measurement error without conic curve approximation when changing liquid material filling condition (d) With conic curve approximation when noise wavelength changes Liquid material minimum part measurement result (e) Liquid material minimum part measurement result with moving average approximation when noise wavelength changes (f) Measurement error α with moving average approximation when noise wavelength changes α Fixed threshold β Individual threshold γ Automatic variation threshold Δ Difference threshold

Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings.
First, FIG. 1 shows a basic configuration of a display panel, particularly a plasma display panel (hereinafter also abbreviated as PDP), which is an object of the present invention. In the PDP 100, a dielectric layer 14 having address electrodes 12 disposed thereon is provided on a rear glass substrate 13, and a partition wall (vertical rib) 11 is provided on the dielectric layer 14, and an RGB phosphor layer 42r, The PDP rear plate 1 to which 42g and 42b are applied, and the front plate 2 having a dielectric layer 22 on which display electrodes 23 are disposed and a protective film 24 are interposed. The discharge space 15 is filled with a mixed gas such as neon or xenon. Here, the light emission principle of the plasma display will be described. For example, when a voltage is applied between the display electrode 23 and a certain address electrode 12a, plasma 101 is generated in the discharge space 15, whereby the phosphor at the selected position emits light, and the display light 102 is transmitted through the front plate 2. Be emitted. A desired color display is performed by a combination of light emission of each phosphor.

  FIG. 2 shows a basic manufacturing flow of the PDP back plate. In the figure, 31 is a cleaning / drying process, 32 is a pattern electrode forming process, 33 is a dielectric layer forming process, 34 is a partition wall forming process, 35 is a phosphor coating process, 36 is a coating process state inspection process, and 37 is a phosphor. The drying process 38 indicates a defect correction process, and the present invention mainly relates to a coating process state inspection process.

  FIG. 3 shows a PDP back plate having no phosphor. Another partition wall (horizontal rib) 16 is formed in the groove partitioned by the partition wall (vertical rib) 11 to form a groove 17 with a horizontal rib. The phosphor is formed at intervals of two for one color along the groove 17 with the lateral rib. Reference numeral 18 denotes one cell surrounded by the vertical rib 11 and the horizontal rib 16. In the figure, the RGB phosphors have different groove widths, but the present invention is not limited to this. Further, the present invention can be applied to a PDP back plate in which the partition walls (lateral ribs) 16 are not formed.

  FIG. 4 shows a state in which the groove 17 with lateral ribs is filled with a liquid material (hereinafter also referred to as a phosphor paste). In particular, when coating is performed while relatively moving an application nozzle having a plurality of holes in the portion corresponding to the groove 17 with the lateral rib to be filled with the phosphor paste 40b in the direction of the coating direction 19, the paste is put into the nozzle holes. When the aggregate, foreign matter, dust, etc. are clogged, the amount of paste applied decreases, resulting in a low filling amount of the phosphor paste 40b ', and finally the paste cannot be applied, resulting in a coating failure 40b ".

  Once the clogging of the coating nozzle occurs, the case of self-healing is very rare, and since defective substrates are continuously manufactured continuously, this clogging defect is detected quickly and the coating process is stopped (in NG substrate manufacturing). Prevention), prompt recovery is the key to improving yield.

  The phosphor paste generally has a relatively high viscosity, and the shape of the surface changes immediately after being filled in the groove 17 with the lateral ribs. Finally, as shown in FIG. A vertical state (both vertical and horizontal ribs) results in a bowl-like shape in which the vicinity is a high portion, and a steady state is obtained. This is called leveling, and the bowl shape is formed as described above as long as the filling amount does not become lower than a certain amount. When the filling amount becomes extremely small or becomes completely zero, the coating omission 41b "occurs.

  After filling the phosphor paste with grooves 17 with horizontal ribs, this is dried to remove the solvent, and as shown in FIG. 6, the phosphor layer is placed at the bottom and sides (vertical ribs, horizontal ribs) of the grooves 17 with horizontal ribs. To cover). Naturally, when the filling amount of the phosphor paste is small, the phosphor layer after drying also becomes thin, resulting in display unevenness when the panel is finally formed. Similarly, the omission of coating causes display defects on the panel.

  A method for measuring the shape of the substrate by the height measuring means is shown in FIG. Missing coating due to nozzle clogging extends from the coating missing point on the substrate to the point where the coating is completed, and continues to occur in the groove 17 with the same lateral rib in the next substrate. That is, it is not necessary to measure the entire surface of the substrate in order to detect nozzle clogging, and all the grooves 17 with the lateral ribs (corresponding to all the holes of the coating nozzle) on the substrate may be inspected across the entire length of the substrate.

  When using a displacement meter 50a with a spot-like measurement area (for example, Keyence, LT8000 series (φ2μm), Keyence, LC series (20 × 30μm), etc.) with all lateral ribs coated with liquid material It is necessary to scan between adjacent partition walls (lateral ribs) 16 like d1 over the entire length of the substrate in the direction crossing the groove 17. Details will be described later.

  When using a displacement meter 50b (for example, OMRON, Z300 series (viewing width 1 mm), KEYENCE, LT9000 series (variable within 2 mm viewing field), etc.) whose measuring area is one-dimensional wide, apply a liquid material. Scan across the entire length of the substrate in the direction across all the lateral groove ribbed grooves 17, but by scanning the distance between adjacent partitions (lateral ribs) 16 like d2 by outputting the average height in the field of view. There is no need. Details will be described later. In FIG. 7, c1 is a transverse rib direction sectional line base point, c1 ′ is a transverse rib direction sectional line end point, c2 is a longitudinal rib direction sectional line (normal application) base point, and c2 ′ is a longitudinal rib direction sectional line (normal application). End point, c3 is a longitudinal rib direction cross-sectional line (abnormal application) base point, c3 ′ is a longitudinal rib direction cross-sectional line (abnormal application) end point, d1 is an example of straight scanning, d2 is an example of oblique scanning, sw1 is a sensor scanning width, sw2 Indicates the sensor field width.

  4 parts (41b) where B phosphor was normally applied, 1 part (41b ') where the application amount decreased due to nozzle clogging, and 1 part (41b') where the application was completely lost due to nozzle clogging ( FIG. 8 shows a state of a cross-section in the same direction (position) as the c1-c1 ′ cross-sectional line of FIG.

  A general displacement sensor operates at a constant response frequency, and when the substrate surface is scanned at a constant speed, the substrate shape data is discretely acquired (sampled). The sampling interval is determined by the sensor response frequency and the sensor scanning speed.

  In order to realize highly accurate measurement, it is preferable to perform as many samplings as possible (in the case of coarse sampling, there is a high possibility of missing the lowest part of the liquid material, which is one of the management indices). For this purpose, (1) There are two methods of using a sensor with a fast response frequency and (2) slowing the scanning speed. However, the specifications of (1) are determined by the sensor manufacturer, and (2) causes an increase in inspection tact, and sufficient sampling cannot be performed within the given inspection tact. Is often considered.

  FIG. 9 shows a sampling timing and a discrete height shape signal 61 obtained when a portion where the coating amount is reduced due to nozzle clogging is measured. The discrete height shape signal 61 is a signal obtained by connecting the height measurement results of the substrate surface obtained by discrete height measurement. As can be seen from FIG. 9, when the lowest part of the liquid material serving as a management index cannot be measured, this is a measurement error (difference from the true value).

  It is known from past experiments that the c1-c1 'cross-sectional shape of the liquid material filled here draws a smooth curve depending on the surface tension of the liquid material and becomes part of a conic curve to reach a steady state. Here, the conic curve is a curve that becomes a boundary of a cross section when the cone is cut along an arbitrary plane, and in the present invention, a circle (a plane that intersects with all the generatrix lines and is parallel to the bottom surface). 4): an ellipse (cut along a plane that intersects all the generatrix lines and is not parallel to the bottom surface), a parabola (cut along a plane parallel to the generatrix line), and a hyperbola (cut along a plane that is not parallel to the generatrix line) Define as type. The c1-c1 'cross-sectional shape of the filled liquid material is the closest to which of these four types of conic curves. Physical conditions such as the three-dimensional shape of the cell and the surface tension and viscosity of the liquid material Therefore, it is preferable to select the most approximate curve and use it for the approximation of the discrete height shape signal. For example, when the surface shape of the liquid material can be approximated by a circle, an approximation including a circular arc based on sampling signals inside (approximate region dw) inside two peak portions (partition wall top) of the discrete height shape signal 61 By calculating a circle and obtaining an approximate curve from the discrete height shape signal, the height shape signal 62 is obtained as shape data close to the actual liquid material surface. The height shape signal 62 is a signal obtained by approximating between discrete height shape signals by an arc of an approximate circle. In FIG. 9, dw is an approximate region, r is an approximate circle radius, PL is a paste surface level, KL is a reference surface level, h is a paste height (PL-KL), and 60a is a discrete height shape outside the approximate region. 60b is an approximate circle, 60d is an approximate circle, 60d is a height shape signal (approximate), 61 is a discrete height shape signal (60a + 60b), and 62 is an approximate circle. As described above, the height shape signal (60a + 60c) is shown.

  FIG. 10 shows a height shape signal 62 and a height signal 63 (a signal obtained by extracting the bottom height of each liquid material from the height shape signal and connecting them to correspond to each liquid material) and an approximate circle radius signal 64 (high From the shape signal, an approximate circle having an arc as the signal portion for each liquid material is obtained, and the diameter of the approximate circle is a signal linked to correspond to each liquid material) and a defect determination threshold (inspection signal is set, The relationship of a threshold for determining the presence or absence of a defect) is shown. FIG. 8A is a cross-sectional view of FIG. 8 with a height shape signal 62 superimposed thereon, and FIG. 8B is a waveform showing only the vertical axis enlarged for easy viewing.

  From the height shape signal 62, the lowest part of the liquid material and the reference surface (for example, the bottom of the groove 17 with the lateral ribs not coated with the liquid material, the glass surface outside the measurement region, etc.) for each of the liquid materials coated with a plurality of lines. The height h is calculated, and as shown in (c), these are connected for each liquid material to obtain a height signal 63. By setting a defect determination threshold value thh as an inspection signal (a signal for determining the presence / absence of a defect with a predetermined threshold value (including both a height signal and an approximate circle radius signal)), the defect portion is set. 63b ′ and 63b ″ are identified.

  The above describes the method for approximating the discrete height signal using a conic curve, taking as an example the case where the surface shape of the liquid material can be approximated by a circle, but the surface shape can be approximated by another conic curve (ellipse, parabola, hyperbola) Is the same. However, when other conic curves (ellipse, parabola, hyperbola) are used, the approximate circle radius signal 64 is naturally not obtained.

  29 and 30, the effect of approximating the discrete height shape signal by the conic curve will be described in detail by taking as an example the case where the c1-c1 ′ cross-sectional shape of the liquid material can be approximated by a circle. As shown in FIG. 29, when the liquid material minimum part cannot be measured because the discrete measurement interval is widened to shorten the time required for the inspection, between the discrete height measurement signal 60b and the actual liquid material minimum part height. Measurement error Em occurs. On the other hand, when the discrete height shape signal is approximated by a circle that is a conic curve and the approximated height shape signal 60d is obtained, the correct minimum height of the liquid material is obtained from the approximated height shape signal 60d. Measurement with little error is possible.

  When the partition wall (vertical rib) 11 with a height of 120 μm is filled with a liquid material in a cell 18 that is arranged so that the distance between the center positions is 350 μm, and the discrete height measurement is performed four times As an example, a theoretical error was calculated for the measurement error with and without conic curve approximation. The results are shown in FIG. In FIG. 30, the horizontal axis represents the height (μm) of the lowest part of the liquid material, and the left vertical axis represents the measured value (μm) corresponding to the conic curve approximation (a) / no approximation (b). The right vertical axis represents the measurement error (μm) corresponding to the measurement error (c) obtained by subtracting the height of the lowest part of the liquid material from the measured value without conic curve approximation (b). Here, the surface shape of the liquid material changes depending on the filling rate of the liquid material itself with respect to the cell capacity. In other words, the liquid material filling rate is high, that is, the measurement error Em without approximating the conic curve becomes smaller because the surface shape of the liquid material approaches the plane as the liquid material minimum part height approaches the partition wall height. The lower the rate, that is, the closer the liquid material minimum part height is to the bottom of the cell, the closer the surface shape of the liquid material is to a circle with a small curvature. This is apparent from the graph of FIG.

  As an actual manufacturing condition, the cell is filled with the liquid material so that the minimum height of the liquid material is 80 to 100 μm. In addition, it has been known that if a failure occurs in the liquid material filling state in the cell, this will lead to a display failure when the panel is commercialized. Must be within ± 10μm. In other words, when an inspection apparatus according to the present invention finds a liquid material filled with a minimum height exceeding the design value ± 10 μm, it is necessary to perform a defect generation process. However, as shown in FIG. 30, when there is no conic curve approximation, a measurement error of 9 to 5 μm occurs, which is not practical. On the other hand, when conic curve approximation is performed, measurement with less error is possible, and defective products can be reliably found and eliminated by high-precision inspection.

  On the other hand, with reference to FIGS. 31 and 32, another effect of approximating the discrete height shape signal by the conic curve will be described in detail by taking as an example the case where the c1-c1 ′ cross-sectional shape of the liquid material can be approximated by a circle. . Here, a case is considered in which, for example, when a roller transporter is used, vibrations generated when the measuring means and the substrate move relative to each other, in particular, vertical movement affects the measurement signal. When vibration occurs in the equipment as described above, the measurement signal itself is also affected by the vibration as shown in FIG. 31, and it is difficult to output a correct measurement result. Therefore, even if the time required for inspection is lengthened, it is common to reduce the discrete measurement interval as much as possible, obtain vibration information along with the shape signal by obtaining a lot of information, and perform processing to remove the vibration element from the measurement signal It is. The noise generated in the measurement signal may include not only the vibration effect of the equipment but also electrical noise received from the amplifier of the measuring machine itself, the power supply of each device, the inverter of the neighboring equipment, and the like. Hereinafter, vibration and electric noise are collectively used as noise for explanation.

  Generally, when noise occurs in a measurement signal, moving average processing is performed on the measurement signal. Moving average processing is a type of frequency filter that is generally used in the field of signal processing, and is an effective technique for removing noise of a specific period from a signal. Specifically, when the wavelength of the noise to be removed is λ, for the Nth signal Y to be processed, the number of signals corresponding to the distance λ is selected together with Y before and after, and the selected signal The values are averaged, and the obtained average value is replaced with the signal value of the Nth signal Y to be processed. Next, similar processing is performed on the (N + 1) th signal Y ', and thereafter, this processing is repeated as needed until the final end of the signal.

  However, when moving average processing is performed, noise having a wavelength equal to or less than the distance to be averaged can be removed, but useful information equal to or less than the distance to be averaged is lost. In other words, when considering the measurement signal in the present invention, if the wavelength λ of noise generated in the measurement signal is larger than the width near the lowest part of the liquid material to be measured, the height information of the lowest part of the liquid material is lost along with noise removal. Will be. Referring to FIG. 31, for example, when the number of discrete measurements on the surface of the liquid material is 13, and the noise wavelength λ corresponds to the length of 5 discrete measurement intervals, a discrete height measurement signal 60b is obtained as the measurement signal. It is done. In this discrete height measurement signal 60b, when noise is removed by moving average processing for five signals, it is clear that even if the noise can be removed, the minimum height of the liquid material is higher than the actual height. In addition, it can be seen that as the average number of signals in the moving average process is increased, the liquid material minimum height is measured to be higher than the actual value. In the example of FIG. 31, the Nth signal value, which is the minimum height of the liquid material, becomes a signal value 60b 'after averaging by the moving average process, and a measurement error Em occurs between the actual liquid material minimum height. On the other hand, let us consider approximating the discrete height shape signal by a circle which is a conical curve. In this case, in consideration of the occurrence of noise in the measurement signal, it is preferable to obtain the height shape signal from the most appropriate approximate circle using as many signals as possible. Furthermore, the measurement signal causing the noise before the approximation is discarded, and the approximate curve is set as the height shape signal 60d. Since the correct minimum height of the liquid material is obtained from the approximated height shape signal 60d, the influence of noise can be eliminated as a result, and accurate measurement is possible.

  A case where the partition wall (vertical rib) 11 having a height of 120 μm is filled with the liquid material in the cell 18 configured so that the distance between the center positions is 350 μm and the minimum part height is 80 μm. As an example, a theoretical value was calculated for the measurement error when the conical curve approximation was performed and when the moving average correction was performed. The results are shown in FIG. In FIG. 32, the horizontal axis represents the noise wavelength as a multiple (times) of the interval between the center positions of the partition walls (vertical ribs) 11, and the left vertical axis represents when conic curve approximation is performed (d) / moving average correction is performed ( The measured value (μm) corresponding to each of e) is taken, and the right vertical axis corresponds to the measurement error (f) obtained by subtracting the height of the lowest part of the liquid material from the measured value when moving average correction is performed (e). Take measurement error (μm). When noise is removed by moving average processing, the measurement error Em is affected by a distance corresponding to the number of signals to be averaged. That is, the shorter the noise wavelength and the shorter the averaging distance, the smaller the difference between the actual liquid material minimum height and the signal value after averaging, so the measurement error Em will be smaller. The longer the distance, the greater the difference between the actual minimum height of the liquid material and the averaged signal value, and the measurement error Em increases. This is apparent from the graph of FIG.

  As described above, as the limit of the quality assurance for filling the liquid material, the actual minimum height with respect to the design value is 10 μm. Here, the noise generated at the actual manufacturing site varies, but if the noise has a wavelength λ that is sufficiently short relative to the width of the lowest part of the liquid material to be measured, the moving average processing is performed by shortening the discrete measurement interval. It is possible to perform measurement with little error. For example, as shown in FIG. 32, if the noise wavelength λ is about 0.15 times the interval between the partition walls (vertical ribs) 11, the measurement error (f) can be suppressed within 1 μm. Withstand practical use. However, when noise at a lower frequency becomes a problem, if noise removal by moving average processing is performed, the measurement error becomes 1 μm or more, and when the noise wavelength λ reaches half of the interval between the partition walls (vertical ribs) 11, The measurement error (f) is 9.5 μm, which is not suitable for practical use. On the other hand, when the conic curve approximation is performed, a measurement error does not occur theoretically, so that highly accurate measurement and inspection can be performed.

  Furthermore, when the c1-c1 'cross-sectional shape of the liquid material can be approximated by a circle, another effect can be expected by using the radius r of the approximate circle as an inspection signal. As described above, the approximate circle radius r obtained when the approximate circle is obtained is calculated for each of the plurality of applied liquid materials, and the approximate circle radius signal 64 is obtained continuously. As the approximate circle radius signal, the approximate circle radius r tends to decrease as the filling amount decreases, but when the coating omission occurs completely, the bottom of the groove 17 with the lateral rib is flat. Because it is, it becomes extremely large. By using the approximate circle radius signal 64 as an inspection signal and setting a defect determination threshold value thr1 (lower side) and thr2 (upper side), 64b ′ and 64b ″ which are signals of the defective part are specified.

  FIG. 11 shows the sensitivity with respect to the filling rate of the phosphor paste as the liquid material (100% when the horizontal ribbed groove 17 is filled) for each of the paste bottom height h and the approximate circle radius r. . In general, in the measurement, it is considered that the higher the change in the evaluation value is, the higher the sensitivity is with respect to the change in the physical quantity (paste filling rate) to be detected. That is, in FIG. 11, it can be said that the higher the slope of the sensitivity characteristic graph, the higher the sensitivity. However, at a low filling rate, the measurement of the paste bottom height h is higher when the paste filling rate is about 80%. It can be seen that the approximate circle radius r measurement is more sensitive. That is, it is preferable to employ a more sensitive measurement according to the manufacturing conditions of the substrate.

  The same direction as the c2-c2 'sectional line and the c3-c3' sectional line in FIG. 7 for the part where the liquid material is normally applied to the groove 17 with the lateral ribs and the part where the filling amount of the liquid material is reduced due to nozzle clogging FIG. 12 shows the cross section of (position). FIG. 12 similarly shows a spot measurement displacement sensor 50a, and the scanning direction of the displacement sensor 50a is from the front side to the back side of the paper.

  As described above, the surface shape of the liquid material becomes bowl-shaped in the cell (space divided by the partition walls (vertical ribs) 11 and the partition walls (lateral ribs) 16) by leveling. However, in this case, the c1-c1 ′ cross-sectional line is shaped like an arc of an approximate circle, but the c2-c2 ′ (c3-c3 ′) cross-sectional line direction is relatively flat at the center, and the partition wall (Horizontal rib) It becomes a shape that becomes a slope in the vicinity of 16.

  Here, regarding the measurement position of the sensor, considering the cell center part p0 and the cell end part p1 of the c2-c2 ′ cross section line of the normal part, the cell center part p0 can accurately measure the height of the lowest part of the liquid material, A higher value than the actual value is output at the cell edge p1. Also, considering the cell center part p0 and cell edge part p1 of the abnormal part c3-c3 'section line, and the p2 on the lateral rib, the height of the lowest part of the liquid material can be accurately measured at the cell center part p0. In the portion p1 and the horizontal rib p2, a value higher than the actual value (≈normal portion height) is output, which causes a defect to be overlooked. Therefore, the sensor scanning needs to be within the sensor scanning width sw1, and it has been experimentally found that the accuracy is preferably within ± 35% with respect to the interval between the partition walls (lateral ribs) 16.

  A specific method for accommodating the sensor scanning width within sw1 will be described. For example, as a substrate moving means 206 for holding a substrate for the actual measurement or moving the substrate to realize sensor scanning, a roller transport machine generally used in the process of manufacturing a glass substrate for display is used. Think about it.

  As shown in FIG. 21, the roller transporter is provided such that a plurality of cylindrical rollers 201 are arranged on a rotating shaft 200 at a predetermined pitch so that the cylindrical side faces in a direction perpendicular to the longitudinal direction of the shaft. The roller shaft 202 has a structure in which a plurality of roller shafts 202 are installed at a predetermined pitch in a direction in which the substrate traveling direction 203 and the roller shaft longitudinal direction are orthogonal to each other. The plurality of roller shafts 202 have their own shaft 200 as a rotation axis. By rotating, the substrate 1d held by the roller 201 is conveyed in the traveling direction 203. Because it is much cheaper than the stage and robot hand using a stone table for the purpose of transporting the substrate in the process, it is frequently used to move the substrate from device to device in the process. However, due to its structure, it is difficult to suppress meandering during movement of the substrate and inclination 205 in the rotation direction about the axis 204 perpendicular to the substrate surface, and vibration generated on the substrate during transport is relatively It is large and generally not used for stroke work other than conveyance such as measurement and processing.

  Also in the case of this measurement, when a spot displacement sensor 50a is installed as shown in the example of FIG. 21 and the substrate 1d is moved by a roller transporter below the sensor to realize scanning of the sensor, Since the inclination 205 in the rotation direction is generated, the scanning locus 51a of the spot displacement sensor 50a cannot be stored in sw1. This is the same when meandering occurs in the movement of the substrate. For the same reason as above, it is also difficult to obtain the relative positional accuracy with respect to the sensor when the substrate is stopped on the roller transporter, and the spot displacement sensor 50a is stopped by a sensor moving mechanism (not shown) after the substrate is stopped. The same problem occurs when the sensor is moved in the sensor moving direction 203 ′ to realize scanning of the sensor.

  In view of the above, as a first method for storing the scanning trajectory 51a of the sensor 50a in sw1, a method using substrate position regulating means as shown in FIGS. 22 and 23 will be described. First, as shown in FIG. 22, when the scanning of the spot displacement sensor 50a is realized by moving the substrate 1d using the substrate moving means 206, the substrate position restricting means 220 is formed at both ends of the substrate edge from the direction perpendicular to the substrate transport direction 203. By limiting the meandering of the substrate and the inclination of the rotation direction about the axis perpendicular to the substrate surface as the central axis, the scanning locus 51a of the spot displacement sensor 50a can be accommodated in sw1.

  As shown in FIG. 23, when the scanning of the spot displacement sensor 50a is realized by moving the sensor by the sensor moving means 231 while the substrate is stopped on the substrate moving means 206, before the sensor scanning, Scanning of the spot displacement sensor 50a is performed by assigning the substrate position regulating means 230 to the four sides of the stopped substrate, and regulating the relative position of the substrate with respect to the sensor and the inclination in the rotation direction about the axis perpendicular to the substrate surface. The locus 51a can be stored in sw1. During the scanning, the position regulating means 230 may remain applied to the substrate or may be separated from the substrate.

  Further, as a second method for keeping the scanning locus 51a of the sensor 50a within sw1, as shown in FIG. 24, scanning for correcting the position of the height measuring means based on the substrate position recognition means and the substrate position information. A method using position correction means will be described. First, as shown in FIG. 24, when the scanning of the spot displacement sensor 50a is realized by moving the substrate 1d using the substrate moving means 206, the substrate position recognizing means 240 measures the edge position of the substrate as needed, and the time elapses. From the accompanying change in the substrate edge position, the meander of the substrate and the inclination in the rotation direction with the axis perpendicular to the substrate surface as the central axis are calculated, and the position of sw1, which is the measurement region, is specified as needed from the obtained information. Using this sw1 position information, the spot displacement sensor 50a can be corrected and moved in the correction direction 203 '' so that the scanning locus 51a of the spot displacement sensor 50a is within sw1 by the scanning position correcting means 241.

  Further, the effect of the above scanning position correction is realized by moving the sensor in the sensor moving direction 203 ′ by the sensor moving means (not shown) while scanning the spot displacement sensor 50a while the substrate is stopped on the substrate moving means 206, and The same can be obtained by moving the substrate position recognizing means 240 by a substrate recognizing means moving means (not shown) in synchronization with the movement of the sensor.

  As a third method for storing the scanning trajectory 51a of the sensor 50a in sw1, as shown in FIG. 25, a method using two or more height measuring means and a sensor interval adjusting means will be described. First, as shown in FIG. 25, two height measuring means are used, and scanning of the first spot displacement sensor 50a and the second spot displacement sensor 50a ′ is realized by moving the substrate 1d using the substrate moving means 206. In this case, the distance adjustment means 250 is used to add the half of the horizontal rib interval to the integral multiple of the horizontal rib interval between the measurement points of the first spot displacement sensor 50a and the second spot displacement sensor 50a ′. The distance between the sensors is preferably set as close as possible so that the housings of the sensor heads do not interfere with each other. When the position is adjusted, even if the board is meandering geometrically or tilted in the rotation direction with the axis perpendicular to the board surface as the central axis, either sensor must be in the center ± 25 mm between the lateral ribs. Within% 25, the scanning locus 51a of the first spot displacement sensor 50a is included in the scanning width sw1 in the region 251a, and the second spot displacement sensor 50a ′ of the region 251b. The scanning trajectory 51a ′ is included in the scanning width sw1 ′, and in the region 251c, the scanning trajectory 51a of the first spot displacement sensor 50a is included in the scanning width w1 ″ ′. In other words, by switching the height measurement means to be referenced depending on the measurement area, even if the substrate meanders or the tilt in the rotation direction about the axis perpendicular to the substrate surface occurs, the information necessary for measurement can be obtained. Can do.

  The effect of using the two sensors is that the scanning of the first spot displacement sensor 50a and the second spot displacement sensor 50a ′ is performed by a moving means (not shown) while the substrate is stopped on the substrate moving means 206. The same effect can be obtained by moving the sensor and the interval adjusting means 250 in the moving direction 203 ′. Further, the number of sensors is not limited to two.

  Up to this point, when the spot measurement sensor 50a is used as the height measurement means, the method for fitting the scanning locus 51a within the measurement width sw1 has been described. In order to ensure the necessary measurement accuracy, the spot measurement displacement sensor There is a method of using a wide displacement sensor 50b instead of 50a.

  As a view similar to FIG. 12, FIG. 13 shows a wide displacement sensor 50b instead of the spot measurement displacement sensor 50a. Similar to FIG. 12, the scanning direction of the displacement sensor 50b is from the front side to the back side.

  The wide displacement sensor 50b outputs an average value of the height in the sensor visual field width sw2, and as shown in FIG. 13, the sensor visual field width sw2 is set to one partition wall (horizontal rib) 16 intervals (between adjacent horizontal rib centers). ), The sensor field of view sw2 always has one cell 18 and one partition wall (even if the sensor measurement position is cell center p0, cell edge p1, and horizontal rib p2). Horizontal ribs) will include 16 widths. Therefore, since there is a difference between the normal part height average value and the abnormal part height average value, it is possible to distinguish these.

  Further, the wide sensor 50b outputs a height profile within the sensor visual field width sw2, and similarly to the above, the sensor visual field width sw2 is set to one partition (horizontal rib) 16 width + partition (lateral rib) 16 interval. The sensor field of view sw2 always includes one cell 18 and one partition wall (lateral rib) 16, even if the sensor measurement position is the cell center p0, cell edge p1, and horizontal rib top p2. It will be. Therefore, the cell center part can be specified from the profile shape, and the measurement result of the cell center part can be output. That is, the data at the cell center p0 can always be acquired, and in principle, the data at the cell edge p1 and the lateral rib p2 is not used for measurement.

  The above is summarized in FIG. FIG. 14 is a result image when the displacement sensor 50 is scanned with coarse accuracy as shown in FIGS. 12 and 13 (oblique scanning d2). When the spot measurement sensor 50a is used, the normal part can be determined to be normal, but when the cell end part p1 and the horizontal rib p2 are scanned with respect to the determination of the abnormal part, an oversight occurs. However, as described above, if the center of the cell can be scanned with high precision (straight forward scanning d1), highly sensitive measurement can be expected, which is more preferable.

  Further, when the average value of the height within the sensor visual field width sw2 is output by the wide-width measuring sensor 50b, it is possible to determine whether the normal part and the abnormal part are normal even in the coarse scanning, and no oversight occurs.

  In addition, when only the height data at the cell center p0 is used for measurement from the height profile within the sensor visual field width sw2 by the wide-width measurement sensor 50b, it is possible to determine whether the normal part and the abnormal part are normal even in the case of coarse precision scanning. So there is no oversight.

  Further, as described above, when a roller transport machine is used for transporting a substrate, for example, vibration generated on the substrate during transport is relatively large as compared with movement on a table using stone. In this inspection, since the shape of the paste is obtained by measuring the height of the substrate surface by the height measuring means, the vertical movement of the substrate surface among vibrations generated in the substrate is directly related to the measurement accuracy.

  As a first method for suppressing the vertical movement of the substrate surface from the measurement data, it is also preferable to provide a substrate back surface measuring means 50c capable of measuring the height of the substrate back surface as shown in FIG. 26A is a view of the measurement system observed from the side with respect to the relative movement directions 203 and 203 ′ of the substrate 1d and the height measuring means 50, and FIG. 26B is a view of the substrate 1d and the height measuring means 50. It is the figure which observed the measuring system from relative movement direction 203 and 203 '.

When the relative movement between the substrate 1d and the height measuring means 50 is realized by moving the substrate 1d in the substrate moving direction 203 using the substrate moving means 206, the fixed height measuring means 50 acquires the surface information of the substrate. However, this surface information includes information on the vertical movement of the substrate. On the other hand, the back surface height measuring means 50c fixed during the same period acquires the back surface information of the substrate, and this back surface information includes only the vertical movement information of the substrate. Therefore, when the substrate back surface information obtained by the back surface height measuring means 50c is subtracted from the substrate surface information obtained by the height measuring means 50, only the substrate vertical movement information is accurately removed from the substrate surface shape. Become. The measurement point of the height measuring means 50 and the measurement point of the back surface height measuring means 50c are preferably positioned at the same point as possible with respect to the substrate plane.

  Further, the above effect is obtained by using the height measuring means moving means 260a and the back surface height measuring means moving means 260b to change the relative movement between the substrate 1d and the height measuring means 50 using the height measuring means 50 and the back surface height measuring means 50c. It can also be obtained when it is realized by moving in synchronization with the sensor movement direction 203 ′. In this case, the vertical movement information of the substrate included in the measurement information of the height measuring means 50 and the back surface height measuring means 50c does not mean that the substrate 1d actually vibrates up and down, and the deflection of the substrate 1d is the main component. Become. In this case, it is necessary to devise the shape of the substrate moving unit so that the back surface height measuring unit 50c can acquire the back surface height information over the entire length in the longitudinal direction of the substrate 1d. For example, as shown in FIG. 26, when a roller transport machine is used as the substrate moving means 206, a space 261 may be provided so that there is no obstacle at the position where the back surface height measuring means 50c scans.

  As a second method for suppressing the vertical movement of the substrate surface from the measurement data, as shown in FIG. 27, the measurement point of the height measurement sensor 50 is installed in a region 262 where the substrate moving means 206 and the substrate 1d are in contact with each other. It is also preferable. 27A is a view of the measurement system observed from the side with respect to the relative movement directions 203 and 203 ′ of the substrate 1d and the height measuring means 50, and FIG. 27B is a view of the substrate 1d and the height measuring means 50. It is the figure which observed the measuring system from relative movement direction 203 and 203 '.

  When the relative movement between the substrate 1d and the height measuring means 50 is realized by moving the substrate 1d in the substrate moving direction 203 using the substrate moving means 206, the fixed height measuring means 50 is fixed to the substrate moving means 206 and the substrate 1d. The surface information can be acquired in a state in which the vertical vibration due to the deflection of the substrate is suppressed by performing measurement in the region 262 in contact with the substrate.

  The above effect is achieved when the relative movement between the substrate 1d and the height measuring means 50 is realized by using the height measuring means moving means 260a to move the height measuring means 50 in synchronization with the sensor moving direction 203 ′. Can also be obtained. In this case, the back surface of the substrate corresponding to the surface of the substrate to be scanned by the height measuring unit needs to be in contact with the substrate moving unit 206 over the entire scanning area. Specifically, the measurement may be performed by leaving the substrate in contact with a high precision table (not shown).

  For the sake of convenience of explanation, the explanation has been given focusing on the use of a roller transport machine as a substrate table when holding the substrate for the actual measurement or moving the substrate to realize the scanning of the sensor. The scope of application of technology is not limited to roller transporters.

  FIG. 15 shows the leveling behavior of the paste. The figure shows the substrate cross-sectional shape (in the same direction as c2-c2 'and c3-c3') immediately after filling the liquid material and after leveling. The liquid material immediately after filling the liquid material is filled not only in the cells 18 but also on the partition walls (lateral ribs) 16. However, the liquid material on the partition wall (lateral rib) 16 flows into the cell 18 according to the paste flow 43. By this action, the surface height of the liquid material is increased immediately after application in the cell central portion p0 as the measurement portion, and the surface height is decreased on the lateral rib p2.

  The relationship between time and leveling is shown in FIG. As shown in FIG. 16, the surface height of the liquid material is increased immediately after application at the cell central portion p0 as the measurement portion, and the surface height is decreased on the lateral rib p2, and this leveling phenomenon is completed and a steady state is obtained. It has been found from the experiment that the time required to reach approximately 5 seconds. However, this is not the case when the viscosity of the liquid material is changed or the design of the substrate is changed, and the leveling behavior needs to be re-evaluated.

  Therefore, it is preferable to wait for 5 seconds until leveling is completed in order to perform highly accurate measurement. This is because if the leveling is not completed at the timing when the cell 18 filled with the liquid material with the normal coating amount is measured, an apparently low numerical value is output, and it is possible that false detection occurs. .

  In order to carry out highly accurate measurement, it is also preferable to perform a correction considering the leveling characteristics as shown in FIG. 16 on the height shape signal measured without waiting for the leveling.

  There are individual differences between coating nozzles and substrates. As the individual difference between the application nozzles, variation in the size of the holes can be considered, and even if the same pressure is applied, the coating amount varies depending on the holes. This characteristic is an individual difference for each nozzle (a certain nozzle always has a large application amount in the X hole, and a certain nozzle always has a small application amount in the Y hole).

  In addition, as the individual difference of the substrate, the variation in cell capacity due to the variation in the width of the partition wall (vertical rib) 11 and the partition wall (lateral rib) 16 due to variations in manufacturing conditions and manufacturing equipment capacity can be considered, and the same filling amount of liquid material is filled. Even so, the height of the liquid material varies depending on the cell capacity. This characteristic is due to individual differences in the substrate due to the manufacturing process (under certain manufacturing conditions, the partition wall width at the substrate end is always narrow and the partition wall width at the center of the substrate is increased).

  What is important here is that the variation in the height of the liquid material due to the individual difference between the coating nozzle and the substrate is not a defect (abnormal coating process) unlike the height variation due to nozzle clogging.

  In general, human vision as a PDP user is said to have a characteristic that distinguishes differences in relative changes with higher sensitivity than differences in absolute changes. That is, when a certain pixel of interest is viewed, if the display luminance there is abruptly lower (higher) than the display luminance of surrounding pixels, it is recognized as display unevenness. On the other hand, even if the display brightness is different from the display brightness of surrounding pixels, it cannot be recognized as display unevenness unless it is large. In other words, liquid material height variations caused by individual differences in the application nozzle and substrate do not surface as product display unevenness, but the change in the liquid material surface height due to nozzle clogging causes the display luminance difference from the surrounding pixels to increase sharply. As a result, the product becomes defective.

  FIG. 17 shows an inspection signal and a fixed threshold value α and a variation threshold value β for defect determination. As for the inspection signal, when there is no defect, signals of s, t, and u are taken. However, it is assumed that t is changed to t ″ and u is changed to u ″ when two defects are generated. When there is no defect, there is a difference in the surface height for each liquid material, but it is a change that gradually increases to the right on the whole, not a product defect. Therefore, for example, although there is a variation in the filling amount due to variation in the hole diameter of the coating nozzle, it is determined that the coating process is operating normally.

  However, when the fixed threshold value α is set in the inspection signal, s which is a normal signal is erroneously detected as a defect. If t ″ and u ″ occur due to the occurrence of two defects, t ″ can be detected but u ″ is overlooked.

  In contrast, if the individual threshold β is set in consideration of the unique hole diameter variation for each coating nozzle and the substrate variation due to manufacturing conditions in advance, t ”and u” can be accurately detected without false detection of s. Can be detected. In FIG. 17, s is, for example, M + 3rd paste height (OK), t is M + 6th paste height (OK), t ″ is M + 6th paste height (NG), u represents the M + 9th paste height (OK), and u ″ represents the M + 9th paste height (NG).

  Also, as shown in FIG. 18, it is preferable to perform an inspection with an automatic variation threshold γ when an inspection signal similar to that in FIG. 17 is obtained, since erroneous detection and oversight are eliminated. The automatic variation threshold γ is used as a threshold value by calculating a moving average signal of the inspection signal itself (a signal obtained by moving average processing of the inspection signal itself).

  Further, as shown in FIGS. 19A to 19E, when determining the substrate to be measured, the time-dependent change amount (difference value) is obtained by referring to the substrate data measured before that time. It is also preferable to perform the inspection by setting the difference threshold value Δ for the amount of change because it prevents false detection and oversight.

  With respect to the inspection signal of FIG. 19, there is no abnormality in the coating process in the Nth and N + 1th coatings, and no defects have occurred in the substrate. The first sheet takes v ′, w ′, and x ′ signals, and there is no significant difference between them. That is, even if the difference threshold Δ is set for the difference value between the N + 1th measurement result and the Nth measurement result shown in FIG. 4D, no defect is detected, which is a normal determination. . However, at the time of coating the N + 2th sheet, two defects occurred in the coating process, and w 'changed to w "and x' changed to x". In this case, a large change occurs in the difference value between the measurement result of the (N + 2) th sheet and the measurement result of the (N + 1) th sheet shown in FIG. (E). If the difference threshold Δ is set, two defects are detected. It can be detected normally. In addition, regarding the reference data that takes a difference from the data to be measured, it is also preferable to use the average of a plurality of data instead of the data of only the previous substrate as described above.

  The above description is based on the assumption that one PDP back plate is manufactured from one glass substrate. However, there are cases where a plurality of PDP rear plates are manufactured from one mother glass substrate 1b for the purpose of tact-up and suppressing the manufacturing cost per substrate.

  As the measurement / inspection technique, the above can be used as it is, but the inspection timing and the target substrate can be selected from the number of lost substrates when NG occurs, the manufacturing tact, the inspection accuracy, and the like. That is, if it is important to reduce the number of lost substrates when NG occurs, the above-described inspection may be performed on all PDP back plates every time the liquid material is applied. If tact-up is aimed at, after applying the liquid material to all the PDP back plates on the mother glass substrate 1b, the above-mentioned inspection may be performed only on the representative substrate, or on the mother glass substrate 1b. After applying the liquid material to all of the PDP back plates, all the substrates may be inspected simultaneously using a plurality of displacement sensors, as will be described later. If the inspection accuracy is important, after applying the liquid material to all the PDP back plates on the mother glass substrate 1b, the height measuring means 50 is applied to the representative substrate only at low speed with high accuracy and low vibration. It is preferable to perform scanning and measurement.

  The purpose of the present invention is to detect abnormalities in the coating process, but because the surface shape of the substrate is measured for this purpose, the surface shape data of all the measured substrates is trend-managed, and the operation of the coating process and nozzles It is also preferable to use it. Specifically, for example, if the surface shape of the liquid material has changed as a whole, a substrate with better quality can be manufactured by adjusting the coating pressure of the coating apparatus. Also, if the surface height of the liquid material is still not abnormal, it can be said that the coating process has not been completed, and an alternative nozzle is prepared as soon as possible. The nozzle can also be replaced before.

  FIG. 20 is a schematic view of an inspection apparatus for realizing the inspection method of the present invention. FIG. 20 shows an example in which six PDP rear plates are manufactured from one mother glass substrate 1a (1b, 1c).

  The liquid material is sequentially applied by the two application means 74 to the mother glass substrate 1b carried in by the substrate carry-in means 75L and fixed on the substrate fixing means 70. For example, application can be performed while moving the application means fixing means 73 with the application means 74 fixed by the moving means 71, and the application of one mother glass substrate 1b is completed by two application operations two by two. . After the coating is completed, the mother glass substrate 1b is unloaded by the substrate unloading means 75UL.

  In order to enable high-precision coating and measurement, the substrate fixing means 70 further has a correction function for the θ direction (rotation direction) with the axis perpendicular to the substrate surface as the central axis in addition to the position correction function for the XY axis. It is also preferable.

  The test is performed on the target substrate at the timing described above after the coating. That is, for example, the height measuring means moving means 72 is moved onto the inspection target substrate by the moving means 71, and the two height measuring means 50 are scanned by the height measuring means moving means 72 to measure the shape of the substrate. . As a result of the inspection, if it is determined that there is an abnormality in the application process, the application process is stopped and a recovery operation is performed. Further, if three height measuring means 50 are provided in the height measuring means fixing means 76, after applying the liquid material to all the PDP back plates on the mother glass substrate 1b, all the substrates are inspected at the same time. It is also possible to do this.

  In the measurement for this inspection on the premise that a plurality of PDP rear plates are manufactured from one mother glass substrate 1b, the height information of the substrate surface is acquired while moving the sensor. For high-accuracy measurement, vertical vibrations during sensor scanning are directly measured and surfaced. For this reason, the sensor scanning mechanism is preferably configured with a mechanism that suppresses vertical movement to the limit. Specifically, an LM guide with an air bearing and a moving mechanism composed of a linear motor can be considered.

  FIG. 28 is a schematic view of another example of an inspection apparatus for realizing the inspection method of the present invention, and shows an example in the case of manufacturing one PDP back plate.

  A fixing unit 280 is provided in a transfer unit that transfers the substrate from the previous process to the next process by the substrate moving unit 206. The fixing unit 280 includes two height measuring units 50a and 50a ′ held by the interval adjusting unit 250. Is provided. The distance between the height measuring means 50a and 50a 'is the shortest distance at which the sensor housing does not interfere with a distance obtained by adding a half length of the transverse rib interval to the integral multiple of the transverse rib interval of the substrate to be manufactured by the interval adjusting means 250 in advance. The distance is adjusted in the distance adjustment direction 203 '' '.

  The substrate 1d is applied with the liquid material on the surface in the previous stroke, and is transported in the substrate transport direction 203 toward the equipment for the next stroke by the substrate moving means 206. At this time, at least over the entire length in the longitudinal direction of the substrate, A part of the surface shape of all the grooves is measured by the height measuring means 50a and 50a ′. As a result of the inspection based on the measurement result, if it is determined that there is an abnormality in the application process, the application process is stopped and a recovery operation is performed.

  Also in the measurement for this inspection, as in the case of another inspection apparatus, the vertical vibration during the substrate scanning becomes a direct measurement error and becomes a surface for high-accuracy measurement. However, in this example, it is preferable to use a general-purpose substrate transport mechanism from the viewpoint of cost and tact, and it is desired to ensure a predetermined inspection accuracy without using a highly accurate stage. Therefore, as shown in FIG. 28, when a roller transport machine is used as the substrate moving means 206, it is also preferable to set the measurement points of the height measuring means 50a and 50a 'on the roller 201 of the roller shaft 202. In FIG. 28, a roller 201 exists immediately below the height measuring means 50a and 50a '.

  In addition to the height measurement means 50a and 50a ′, as another method for eliminating the influence of vertical vibration during substrate scanning from the measurement data when the general-purpose substrate transfer means 206 is used, the substrate back surface height measurement means It is also preferred to use 50c and 50c ′. A sensor installation example of the inspection apparatus according to the present case is shown by a dotted line in FIG. That is, if the substrate back surface information obtained by the back surface height measuring means 50c and 50c ′ is subtracted from the substrate surface information obtained by the height measuring means 50a and 50a ′, only the substrate vertical movement information is accurately obtained from the substrate surface shape. Will be removed. Note that the measurement points of the height measuring means 50 and 50a 'and the corresponding measurement points of the back surface height measuring means 50c and 50c' are preferably positioned at the same point as possible with respect to the substrate plane.

  In addition, a substrate that has become NG due to a defect in the coating process can be restored as a non-defective product by correcting it with a dispenser that can be manually filled with a liquid material.

  Examples of the present invention are specifically shown below. However, the content of the present invention is not limited to this.

Example 1
The PDP back plate to be measured is the partition 18 (vertical rib) 11 shown in FIG. 3 divided by the partition (transverse rib) 16 to form cells 18, each set of RGB cells having different groove widths. Thus, one pixel of the PDP is formed. The width of the cell 18 divided by the partition walls (lateral ribs) 16 is 950 μm, and the width of the partition walls (lateral ribs) 16 is 50 μm (the interval between the partition walls (lateral ribs) 16 is 1000 μm). In addition, it is assumed that six PDP rear plates 1b1 to 1b6 are positioned on the mother glass substrate 1b in a height measurement scanning direction of 2 sheets × phosphor application direction of 3 sheets. The liquid material filled in the groove 17 with the lateral ribs is a phosphor paste in which phosphor materials that promote the color development of each RGB are dissolved in a solvent. In the first embodiment, the substrate is not composed of an RG phosphor. Consider the case where the B phosphor paste 40b is applied at a filling amount of 75% of the cell capacity.

  The apparatus shown in FIG. 20 is used as an apparatus for applying the phosphor paste and an apparatus for inspecting the state of the application apparatus. First, regarding the phosphor paste coating function, the coating means 74 uses two coating nozzles in which a plurality of nozzle holes are arranged one-dimensionally at positions corresponding to the plurality of lateral rib grooves 17 to which the phosphor is to be coated. . Further, as the application means fixing means 73 for fixing the application nozzle and the moving means 71 for moving the application means fixing means 73 in the coating direction 19, a gantry stage having a positioning / correction function on the XYZ axes is used.

  Next, regarding the inspection function, two triangulation laser displacement meters LC-2430 (manufactured by Keyence Corporation) having a spot measurement field are used as the height measuring means 50. Also, the height measuring means moving means 72 and the moving means 71 for positioning the height measuring means 72 on the substrate to be measured (used in common with the coating gantry) have positioning / correction functions on the XYZ axes, respectively, and scanning In order to suppress the inner vibration as much as possible, we decided to use an LM guide and a gantry stage composed of a linear motor with air bearings. The apparatus was designed such that the scanning linear advance ability of the laser displacement meter 50 by the LM guide 72 is the cell center position ± 300 μm. When the inspection function is configured as described above, the three height measuring means 50 and the height measuring means fixing means 76 shown above the substrate carry-out means 75UL are not necessarily required.

  In addition, a general-purpose high-precision stage was used as the substrate fixing means 70 for positioning the mother glass substrate 1b with high accuracy and correcting the positions of the XY and θ axes. Furthermore, general-purpose roller transport mechanisms were used as the substrate carry-in means 75L and the substrate carry-out means 75UL for carrying in and out the mother glass substrate 1a (1b, 1c) into the apparatus.

  The operation of the coating function and the loading and unloading of the mother glass substrate 1a (1b, 1c) into the apparatus, the movement of the inspection function, and the scanning were performed intensively by the coating apparatus operation unit 78, and obtained by the height measuring means. The electrical signal processing is performed by the inspection device operation unit 77, and the application device operation unit 78 and the inspection device operation unit 77 are electrically controlled by a general-purpose PLC (not shown) so that they can communicate with each other. ing. The inspection device operation unit 77 further includes an input / output device such as a general-purpose personal computer (not shown) that performs signal processing, a keyboard that serves as an interface with the operator, a mouse, and a monitor that outputs measurement results and inspection results. .

Hereinafter, description will be made while following the operations of the coating apparatus and the inspection apparatus.
First, the mother glass substrate 1b carried onto the high precision stage 70 by the roller transport mechanism 75L is fixed on the high precision stage 70 by vacuum suction or the like, and then the XY and θ axes are finely adjusted to be in a predetermined position. Positioned. Next, the phosphor coating nozzle 74 is positioned by the gantry 73 and the gantry stage 71 at the coating start position (for example, the XDP origin direction end of the PDP back plates 1b1 and 1b2) and finely adjusted in the XYZ axis directions. The phosphor coating nozzle 74 positioned at the coating start position applies the phosphor paste, which is the coating liquid, to the groove 17 with the lateral ribs on the substrate surface by pressurizing the inside of the nozzle, and this operation is completed for the gantry stage 71. The application of the phosphor paste to a predetermined position over the entire length of the substrate is completed by continuously performing the movement toward the position (for example, the opposite ends of the PDP rear plates 1b1 and 1b2 in the X-axis origin direction). The mother glass substrate 1b in FIG. 20 has been applied with the phosphor paste on the PDP back plates 1b1 and 1b2 by the coating operation, and the PDP back plates 1b3 to 1b6 are in a stage where the phosphor paste is not applied.

  By repeating the above coating operation three times sequentially, the coating of the phosphor paste is completed on all of the PDP back plates 1b1 to 1b6, and the mother glass substrate 1c is carried out to the subsequent process by the roller transport machine 75UL.

  In Example 1, in order to minimize the number of lost NG substrates produced when the coating nozzle is clogged, all substrates are measured and inspected every time coating is performed. That is, the outline of the operation is as follows: loading of the mother glass substrate 1b → application of the PDP rear plates 1b1, 1b2 → inspection of the PDP rear plates 1b1, 1b2 → application of the PDP rear plates 1b3, 1b4 → inspection of the PDP rear plates 1b3, 1b4 → PDP back plate 1b5, 1b6 application → PDP back plate 1b5, 1b6 inspection → mother glass substrate 1b unloading.

  As described above, the substrate manufacturing conditions are set so that the phosphor paste B is applied at a filling amount of 75% of the cell capacity, so the discrete height shape signal is one of the conic curves as the inspection signal. Using the height signal obtained from the height shape signal obtained by approximating with an approximation method using a parabola, the defect determination threshold thh is an individual individually adjusted manually considering the hole diameter variation of the application nozzle A threshold β was applied. In addition, in order to prevent a decrease in inspection accuracy due to the paste leveling operation, the laser displacement meter was scanned over the portion where 5 seconds had elapsed since application.

  As a result, in the coating process, the phosphor paste has been applied to the substrate smoothly from the start of manufacture. However, in a certain period of time, of the two coating nozzles, the M1th hole of the coating nozzle on the substrate carry-out side was clogged with dust mixed in the nozzles during nozzle assembly, and the coating amount decreased. As a result, the filling amount of the phosphor paste in the M1th lateral ribbed groove 17 corresponding to the M1th hole is reduced from about 75% to about 70%, and the surface height h of the lowest part of the paste is about 75 μm to 65 μm. Decreased back and forth. In another time zone, the M2th hole of the coating nozzle on the substrate carry-in side of the two coating nozzles is clogged with dust mixed in the phosphor paste during manufacture of the phosphor paste and cannot be completely coated. It became. As a result, in the M2th groove 17 with the lateral rib corresponding to the M2th hole, the phosphor paste was missed. The inspection apparatus detected these normally, and once the coating apparatus was stopped and the coating nozzle was replaced quickly, the process could be quickly restored to normal with the minimum number of NG substrate losses. In addition, while the coating was performed smoothly, no false detection / overdetection by the inspection device occurred.

Example 2
In the form of Example 1, the substrate manufacturing conditions were reset so that the phosphor paste B was applied at a filling amount of 90% with respect to the cell capacity. In response to this, an approximate circle radius signal obtained using a circle that is one of conic curves is used as an inspection signal to approximate the discrete height shape signal, and the defect determination threshold values thr1 and thr2 are centered from both ends of the substrate. Taking into account the manufacturing condition of the substrate due to the drying furnace characteristics that the partition walls (vertical ribs) 11 tend to become thicker in the vicinity, the moving average signal of the inspection signal itself is automatically obtained, and the fluctuation threshold γ adjusted based on this is obtained. Applied. In order to shorten the inspection tact, the laser displacement meter started scanning from the portion 2 seconds after application, and the measurement was performed. In order to prevent a decrease in inspection accuracy due to the paste leveling operation, the relationship of FIG. The test was carried out after correcting the height shape signal.

  As a result, in the coating process, the phosphor paste has been applied to the substrate smoothly from the start of manufacture. However, in a certain period of time, among the two coating nozzles, the M3th hole of the coating nozzle on the substrate carry-in side was clogged with dust mixed in the phosphor paste during the production of the phosphor paste, and the coating amount decreased. As a result, the filling amount of the phosphor paste in the M3th lateral ribbed groove 17 corresponding to the M3th hole is reduced from about 90% to about 85%, and the approximate circular radius r of the paste surface shape is about 400 μm to about 270 μm. Decreased back and forth. The inspection device detected this normally, and once the coating device was stopped and the coating nozzle was quickly cleaned, the process could be quickly restored to normal with the minimum number of NG substrate losses. In addition, while the coating was performed smoothly, no false detection / overdetection by the inspection device occurred.

Example 3
In the form of Example 2, the inspection apparatus was remodeled for further substrate manufacturing tact time up. A sufficiently rigid frame is installed as the height measuring means fixing means 76 on the upper part of the substrate unloading means 75UL, and the height measuring means 50 has a width of 1000 μm which is the same as the width of the cell 18 and one partition wall (lateral rib) 16. Three wide laser displacement meters 50b having a measurement area and set to output the average height in the measurement area were installed. As this wide laser displacement meter, for example, a triangulation laser shape measurement sensor Z300-S10 (OMRON) having a measurement visual field can be used. When the inspection function is configured as described above, the two laser displacement meters and the gantry stage 72 having a spot field as the height measuring means 50 used in the first and second embodiments are not necessarily required.

  In Example 3, for the substrate manufacturing tact-up as described above, the mother glass substrate 1b was unloaded and fixed upward while the application of all the PDP back plates 1b1 to 1b6 on the mother glass substrate 1b was completed. The substrate surfaces of all the substrates are measured with a laser displacement meter, and inspection is performed. That is, the outline of the operation is as follows: loading of the mother glass substrate 1b → application of the PDP rear plates 1b1, 1b2 → application of the PDP rear plates 1b3, 1b4 → application of the PDP rear plates 1b5, 1b6 → unloading the mother glass substrate 1b → PDP This is an inspection of the back plates 1b1 to 1b6. With the configuration of the third embodiment, it is possible to apply the phosphor paste to all the PDP rear plates without waiting for the time for inspection, and it becomes unnecessary to precisely position the sensor scanning locus by the gantry stage, and Since all the substrates can be inspected together at the same time, the substrate manufacturing tact time can be greatly reduced.

  In addition, as the defect determination thresholds thr1 and thr2, considering the manufacturing state of the substrate due to the drying furnace characteristics that the partition walls (vertical ribs) 11 tend to gradually increase from both ends of the substrate to the vicinity of the center, The inspection signal average value of 10 substrates before the measurement was obtained, and the difference threshold value Δ was applied to the difference value between the inspection signal average value and the inspection signal obtained from the substrate to be inspected, and the inspection was performed.

  As a result, in the coating process, the phosphor paste has been applied to the substrate smoothly from the start of manufacture. However, in a certain period of time, among the two coating nozzles, the M4th hole of the coating nozzle on the substrate carry-out side was clogged with aggregates generated in the phosphor paste itself, and the coating amount was reduced. As a result, the filling amount of the phosphor paste in the M4th lateral rib groove 17 corresponding to the M4th hole is reduced from about 90% to about 80%, and the approximate circular radius r of the paste surface shape is about 400 μm to 210 μm. Decreased back and forth. The inspection apparatus detected this normally, and once the coating apparatus was stopped and the coating nozzle was washed quickly, the process could be quickly restored to normal. In addition, while the coating was performed smoothly, no false detection / overdetection by the inspection device occurred.

Example 4
In the form of Example 1 above, when the surface shape data of the substrate was temporally managed and compared in the order of measurement and inspection, the entire fluorescence was observed within the substrate surface at the timing of phosphor paste lot switching. It was found that the height h of the lowest part of the body paste was increased. The cause was determined to be a decrease in paste viscosity due to phosphor paste manufacturing variation, and when the coating pressure of the coating machine was adjusted, the state before the lot change was restored. Further, it was found that the surface height h of the lowest portion of the paste is decreasing in the M5th lateral rib groove 17 corresponding to the M5th hole of the coating nozzle on the substrate carry-in side of the two coating nozzles. On the other hand, it was possible to prevent clogging of the coating nozzles by cleaning the nozzle holes early.

Example 5
In the form of Example 3 above, while the phosphor paste is being applied to the PDP back plate 1b2 on a certain mother glass substrate 1b, the fluorescent light is produced in the M6th hole of the coating nozzle on the substrate carry-out side when the phosphor paste is manufactured. Garbage mixed in the body paste was clogged and the coating amount decreased. The inspection apparatus detected this normally, but the PDP rear plates 1b2, 1b4, and 1b6 on the substrate carry-out side on the mother glass substrate 1b became NG substrates. However, this NG substrate was pulled out from the process, fixed on a table for correction, and corrected with a dispenser that can be manually filled with a liquid material.

Example 6
The PDP back plate to be measured is the partition 18 (vertical rib) 11 shown in FIG. 3 divided by the partition (transverse rib) 16 to form cells 18, each set of RGB cells having different groove widths. Thus, one pixel of the PDP is formed. The width of the cell 18 divided by the partition walls (lateral ribs) 16 is 900 μm, and the width of the partition walls (lateral ribs) 16 is 50 μm. In the sixth embodiment, it is assumed that one PDP back plate 1d is produced for one glass substrate. The liquid material 40b filled in the groove 17 with the lateral rib is a phosphor paste in which a phosphor material that promotes the color development of each RGB is dissolved in a solvent. In the sixth embodiment, the liquid material 40b is applied to a substrate on which no RG phosphor is formed. Consider the case where B phosphor paste is applied at a filling amount of 75% of the cell capacity.

  A coating device (not shown) is used as a device for applying the phosphor paste, and a device shown in FIG. 28 is used as a device for inspecting the state of the coating device.

  In the inspection apparatus, as the height measuring means 50, two triangulation laser displacement meters LK-G10 (Keyence) having a spot measurement field are used (spot displacement sensors 50a and 50a '). A general-purpose automatic single-axis stage is used as the distance adjusting means 250 for adjusting the distance between the spot displacement sensors 50a and 50a ', and is attached to a sensor frame as the fixing means 280. Further, the field of view of the spot displacement sensors 50a and 50a ′ is set on a roller 201 which is a part of a roller transport machine as the substrate transport means 206, and the measurement point interval of both sensors is adjusted in advance before performing the inspection. It is set to 63450 μm by means 250.

  Processing of the electrical signal obtained by the height measuring means 50 is performed by the inspection apparatus operation unit 281. The inspection apparatus operation unit 281 further includes an input / output device such as a general-purpose personal computer (not shown) that performs signal processing, a keyboard that serves as an interface with an operator, a mouse, and a monitor that outputs measurement results and inspection results. .

Hereinafter, description will be made while following operations of a coating apparatus and an inspection apparatus (not shown).
First, the phosphor is applied to the substrate surface by a coating machine (not shown). As a coating machine, a nozzle coating type coating machine having a mechanism as described in Example 1 is conceivable. However, as described above, in Example 6, a single-sheet substrate is assumed. It is a specification for the corresponding single sheet application. Specifically, it has only one coating nozzle and discharges the substrate every time the coating operation for one sheet is completed, and carries in a new substrate.

  When the application of the phosphor paste to the substrate surface by the coating machine is completed, the substrate 1d is unloaded to the subsequent process by the roller transporting machine 206. On the roller transporting machine, the surface shape of the substrate being transported is captured. Is provided with a height measuring means 50.

  As described above, the substrate manufacturing conditions are set so that the phosphor paste B is applied at a filling amount of 75% of the cell capacity, so the discrete height shape signal is one of the conic curves as the inspection signal. The height signal obtained from the height shape signal obtained by the approximation method using a parabola is used, and the defect determination threshold thh is determined by taking into account the hole diameter variation of the coating nozzle and manually adjusting the individual threshold β individually adjusted. Applied. Further, the signal obtained when the partition (lateral rib) 16 is scanned is excluded from the signals obtained from the two spot displacement sensors 50a and 50a ′, and either of the spot displacement sensors 50a and 50a ′ is within the scanning width. Implemented signal processing to extract and link signals when scanning.

  As a result, during substrate transfer, a meandering of ± 400 μm and ± 400 μm meandering with respect to the substrate transfer direction 203 occurred in the rotation direction with respect to the axis perpendicular to the substrate surface, but from the spot displacement sensors 50a and 50a ′ By extracting the signal when scanning within the scanning width and obtaining a continuous signal, the signal when scanning on the partition wall (lateral rib) 16 is eliminated, and all grooves can be measured with certain accuracy. confirmed.

  In the coating process, the phosphor paste has been applied to the substrate smoothly since the start of production. However, in a certain period of time, the M7th hole of the coating nozzle was clogged with dust mixed in the nozzle during nozzle assembly, resulting in a decrease in the coating amount. As a result, the filling amount of the phosphor paste in the M7th lateral ribbed groove 17 corresponding to the M7th hole is reduced from about 75% to about 60%, and the surface height h of the lowest part of the paste is about 75 μm to 32 μm. Decreased back and forth. The inspection apparatus detected these normally, and once the coating apparatus was stopped and the coating nozzle was replaced quickly, the process could be quickly restored to normal with the minimum number of NG substrate losses. In addition, while the coating was performed smoothly, there was no false detection / overdetection by the inspection machine.

Example 7
In the embodiment 6, the measurement visual fields of the spot displacement sensors 50a and 50a ′ are removed from the roller 201 which is a part of the roller transport machine as the substrate transport means 206, and two substrate back surface height measuring means 50c provided further. And the configuration was changed so that measurement could be performed with the substrate sandwiched at 50c '. Also, signal processing is performed to eliminate the substrate vertical movement signal included in the measurement signal by subtracting the measurement signal obtained by the back surface height measuring means 50c (50c ′) from the measurement signal of the spot displacement sensor 50a (50a ′). added.

  As a result, while the coating machine is applying the phosphor paste to the substrate 1d, the M8th hole of the coating nozzle is clogged with dust mixed in the phosphor paste during manufacturing of the phosphor paste, and the coating amount is around 75%. The surface height h of the lowest part of the paste decreased from around 75 μm to around 65 μm. The inspection device detected this normally, and once the coating device was stopped and the coating nozzle was quickly cleaned, the process could be quickly restored to normal with the minimum number of NG substrate losses. In addition, while the coating was performed smoothly, there was no false detection / overdetection by the inspection machine.

  As mentioned above, about Examples 1-7, although the case where B phosphor paste is apply | coated to the board | substrate with which RG phosphor is not comprised is considered, this is not limited to B phosphor paste, For convenience, a phosphor layer having a color other than the phosphor paste to be measured may be formed in another laterally ribbed groove 17. In this case, the height h may be calculated on the basis of the height outside the region where the lateral ribbed grooves 17 serving as the measurement region exist, for example, the raw glass surface height. Further, since the height of the partition walls (vertical ribs) 11 is known from the substrate design value, the height h can be calculated based on the height of the partition walls (vertical ribs) 11.

Claims (36)

  The height of the substrate surface including the liquid material application part while having the height measurement means and moving the substrate or the height measurement means in a direction intersecting with the liquid material applied to the substrate at a predetermined interval. Measurement signals are discretely obtained, and an approximate curve is obtained from the obtained discrete height shape signal. The height signal for each liquid material extracted from the height shape signal obtained from the height shape signal is used as the inspection signal. A method for inspecting a display panel, characterized in that the coating amount for each liquid material is measured.   The liquid material application part signal is identified from the discrete height shape signal obtained by the height measuring means, and the height shape signal is obtained from the identified signal using a conic curve as an approximate curve. Item 6. A display panel inspection method according to Item 1.   The liquid material application part signal is identified from the discrete height shape signal obtained by the height measuring means, and the height shape signal is obtained from the identified signal using a circle as an approximate curve, and the diameter of the approximate circle is plural. 3. The display panel inspection method according to claim 2, wherein an approximate circle diameter signal linked to correspond to the liquid material is used as an inspection signal, and an application amount for each liquid material is measured from the inspection signal.   On the substrate, a plurality of first partition walls are formed at predetermined intervals in a direction parallel to the longitudinal direction of the liquid material applied at predetermined intervals, and the liquid material is further interposed between adjacent first partition walls. 4. The display panel inspection method according to claim 1, wherein a plurality of second partitions are formed at predetermined intervals in a direction perpendicular to the longitudinal direction of the display panel. 5.   Using a height measurement sensor having a spot-like measurement region as the height measurement means, the shape of the region within ± 35% of the central portion between the second partition walls formed in the direction perpendicular to the longitudinal direction of the liquid material is 5. The display panel inspection method according to claim 4, wherein measurement is performed over the entire length of the substrate in a direction crossing the longitudinal direction of the liquid material.   A substrate position regulating means for regulating the position of the substrate is further provided, and the shape of the region within ± 35% of the central portion between the second partition walls is measured over the entire length of the substrate in the direction crossing the longitudinal direction of the liquid material. The display panel inspection method according to claim 5.   A substrate position recognizing means for recognizing the position of the substrate and a scanning position correcting means for correcting the position of the height measuring means based on the substrate position information are further provided, and the area within ± 35% of the central portion between the second partition walls 6. The display panel inspection method according to claim 5, wherein the shape is measured over the entire length of the substrate in a direction crossing the longitudinal direction of the liquid material.   It further has two or more height measuring means, position adjusting means, and switching means, and the shape of the region within ± 35% of the central portion between the second partition walls is formed in the direction across the longitudinal direction of the liquid material. 6. The method for inspecting a display panel according to claim 5, wherein measurement is performed over a range.   A height measurement sensor having a measurement region including a second partition wall interval formed in a direction perpendicular to the longitudinal direction of the liquid material is used as the height measurement means, and the shape of the substrate surface is set in a direction crossing the longitudinal direction of the liquid material. 5. The display panel inspection method according to claim 4, wherein measurement is performed over the entire length of the substrate.   The substrate back surface height measuring means for measuring the height of the substrate back surface is provided, and the measurement result by the height measuring means is corrected by the substrate back surface height measurement result. The inspection method of the display panel as described.   The display panel inspection method according to claim 1, wherein the measurement position of the height measuring unit is arranged at a position where the substrate moving unit and the substrate are in contact with each other.   The liquid material applied at a predetermined interval changes the surface shape between the first and second partition walls by a fluid action immediately after application, and reaches a steady state after a predetermined time. The display panel inspection method according to claim 1, wherein the display panel inspection method is performed after a predetermined time.   The liquid material applied at a predetermined interval changes the surface shape between the first and second partition walls by a fluid action immediately after application, and reaches a steady state after a predetermined time. 12. The method for inspecting a display panel according to claim 1, wherein the height shape signal is corrected based on the change information.   In the signal processing step of providing a predetermined defect determination threshold for determining the presence or absence of a defect in the inspection signal, each region corresponding to a plurality of liquid materials to be measured in the inspection signal is specified, and each of the specified signal portions 14. The display panel inspection method according to claim 1, wherein a unique defect determination threshold is provided.   15. The display according to claim 14, wherein a defect determination threshold for the inspection signal is automatically adjusted based on a moving average signal obtained by performing a moving average process on the inspection signal itself obtained from the inspection target substrate. Panel inspection method.   The board surface height is measured continuously for multiple boards, and the defect judgment threshold of the board to be inspected is automatically determined from the board height and shape information measured before the measurement of the board to be inspected. The display panel inspection method according to claim 14, wherein adjustment is performed.   Every time the liquid material is applied to the substrate, the height measurement is performed on all the substrates to which the liquid material is applied, or all the liquid materials are applied after the liquid material is applied to a plurality of substrates. The display panel inspection method according to claim 1, wherein the display panel inspection method is performed on a substrate or a selected representative substrate.   18. The display panel inspection method according to claim 17, wherein height measurement information obtained from a plurality of substrates is managed and fed back to control and operation of the coating apparatus.   A height measuring means for discretely measuring the height of the substrate surface including the liquid material application part, and a signal processing means for obtaining an approximate curve from the obtained discrete height shape signal to obtain a height shape signal. A display panel inspection device.   A moving means for moving the substrate or the height measuring means in a direction intersecting with the liquid material applied to the substrate at a predetermined interval; and an information output means for outputting the measurement result and the inspection result by the signal processing means. 20. The display panel inspection apparatus according to claim 19, further comprising:   21. The display panel according to claim 20, further comprising substrate fixing means for fixing the substrate, wherein the substrate fixing means has a position correcting function in the rotation direction about an axis perpendicular to the substrate surface. Inspection equipment.   A laser displacement meter is used as the height measuring means, a linear motor guide equipped with an air bearing is used as the moving means for moving the height measuring means, and the axis perpendicular to the substrate surface is used as the central axis as the substrate fixing means for fixing the substrate. The display panel inspection apparatus according to claim 21, wherein the display panel inspection apparatus is configured using a high-precision stage having a position correction function in a rotational direction.   23. The display panel inspection apparatus according to claim 22, wherein a high-precision stage as a substrate fixing means for fixing the substrate is used in common with the coating apparatus as the substrate fixing means when the liquid material is applied. .   21. The display panel inspection apparatus according to claim 20, further comprising substrate position regulating means for regulating the position of the substrate.   The laser displacement meter is used as the height measuring means, a roller transport machine is used as the moving means for moving the substrate, and a position restricting guide is used as the substrate position restricting means. Display panel inspection equipment.   The laser displacement meter is used as the height measuring means, the single-axis stage is used as the moving means for moving the height measuring means, and the positioning mechanism is used as the substrate position restricting means. The display panel inspection apparatus as described.   21. The display panel inspection apparatus according to claim 20, further comprising position correction means for correcting the positions of the substrate edge position measurement means and the height measurement means.   A laser displacement meter is used as the height measuring means, a roller transport machine is used as the moving means for moving the substrate, a laser position measuring sensor is used as the substrate edge position measuring means, and a single-axis stage is used as the position correcting means. 28. The display panel inspection apparatus according to claim 27, wherein:   21. The display panel inspection apparatus according to claim 20, further comprising an installation interval adjusting unit that adjusts an installation interval between at least two height measuring units and the height measuring units.   30. The apparatus according to claim 29, wherein two laser displacement meters are used as the height measuring means, a roller transporter is used as the moving means for moving the substrate, and a single-axis stage is used as the installation interval adjusting means. The display panel inspection apparatus according to 1.   31. The display panel inspection apparatus according to claim 24, further comprising a substrate back surface height measuring unit, wherein a laser displacement meter is used as the substrate back surface height measuring unit.   31. The display panel inspection apparatus according to claim 24, wherein a laser displacement meter as a height measuring unit is configured to measure a position where the substrate moving unit and the substrate are in contact with each other.   A display panel manufacturing method using the inspection method according to claim 1 to manufacture a display panel.   34. The method of manufacturing a display panel according to claim 33, wherein the substrate is corrected using a liquid material correcting means based on the defect information of the substrate.   A display panel manufacturing method using the inspection apparatus according to claim 19 to manufacture a display panel.   36. The method for manufacturing a display panel according to claim 35, wherein the substrate is corrected by using a liquid material correcting means based on the defect information of the substrate.

Images