JP2967618B2 - Blade position detector - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a blade position detecting device, and more particularly to a blade position detecting device for detecting breakage, wear, and the like of a high-speed rotating blade of a dicing device for performing a groove cutting process on a wafer in semiconductor manufacturing or the like. .

[0002]

2. Description of the Related Art It is an important factor to make the uncut amount of a workpiece equal to a set value with a dicing apparatus. It is necessary to perform the correction with high accuracy and detect and correct the wear amount of the blade.

[0003] The following two methods are known as methods for correcting the blade wear amount. The first method is a method in which, after processing a predetermined number of processing lines of a workpiece, the blade is brought into contact with the processing table to energize the blade and the processing table, thereby correcting the blade. Second
Is a method in which blade wear amount data (experience value) is input to a dicing apparatus in advance as an automatic wear correction amount.

[0004]

However, the first method has a problem that the blade is damaged because the blade is brought into contact with the processing table, and furthermore, the processing table is damaged. In addition, the second method has a problem that the amount of wear of the blade varies depending on conditions such as the type of a workpiece to be machined, the variation of the blade, and the amount of cutting.

[0005] The present invention has been made in view of such circumstances, and without causing the blade to contact the processing table, the amount of wear of the blade depends on conditions such as the type of work to be processed, blade variation, and the amount of cutting. It is an object of the present invention to propose a blade position detecting device capable of positioning the tip of the blade with high accuracy in response to the change.

[0006]

According to the present invention, in order to achieve the above object, the present invention is provided so as to be opposed to a light projecting means at a predetermined interval, and the light projected from the light projecting means is provided on one side. A pair of optical systems for condensing the light condensed within the predetermined interval by the optical system, and condensing the light diffused after condensing with the other optical system, and receiving the light condensed by the other optical system for photoelectric conversion A light receiving means, a rotating blade that moves orthogonally to the optical axis between predetermined intervals of the optical system, and blocks light condensed within the predetermined distance, and reads the amount of light received by the light receiving means and rotates the light receiving means. Means for reading the displacement of the blade; and means for outputting a signal when the amount of light received by the light receiving means matches a preset value stored in advance. A blade position detection device that positions the tip of the rotary blade with respect to a workpiece, A blade position detecting device, wherein the blade position detecting device is provided apart from the processing table of the rotary blade, the cleaning device comprising: a cleaning unit for jetting water and air to one of the pair of optical systems and the other facing surface; The means ejects water during processing to form a water film on one and the other opposing surfaces of the optical system, and ejects air at the time of detecting the position of the rotary blade to attach water droplets on one and the other opposing surfaces of the optical system. Is removed.

[0007]

According to the present invention, the cleaning means can jet water and air to the opposing surfaces of one and the other of the pair of optical systems. A water film is formed on the facing surface, and air can be blown out when the position of the rotary blade is detected to remove water droplets attached to one and the other facing surfaces of the optical system.

[0008] Therefore, mist containing chips generated during work processing can be prevented from adhering to the opposing surface of the optical system, so that the opposing surface of the optical system is kept clean and guided to the light receiving means. The light amount can be kept constant.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of a blade position detecting device according to the present invention will be described below in detail with reference to the accompanying drawings. As shown in FIG. 1, the clock 12 of the blade position detecting device 10 drives the transistor 14 to turn the light emitting diode 16 to “O”.
N "and at the same time, the first switch 18 is turned" OF ".
F ", the second switch 20 is turned" ON ". The clock 12 drives the transistor 14 to turn the light emitting diode 16 to the “OFF” state, and at the same time, turns the first switch 18 “ON” and the second switch 20 “OF”.
F ”. Thereby, the operational amplifier 2 described later
2 is connected to the amplifier 24 and to the amplifier 26.

The optical fiber 28 has a light emitting diode 1 at one end.
6 and the other end is disposed on the inspection table 30. The inspection table 30 is provided on the base 31,
The inspection table 30 is provided with a lens system 34 so as to be located above the other end of the optical fiber 28. Further, prisms 36 and 38 are provided on the table of the inspection table 30.

The prism 36 is located above the lens system 34, and a lens system 40 is provided below the prism 38. The optical fiber 32 is located below the lens system 40.
Is provided, and the other end is provided with the operational amplifier 22 described above.
In the vicinity of the photodiode 23. Therefore, the light projected from the light emitting diode 16 enters the lens system 34 via the optical fiber 28, and the light refracted by the lens system 34 is reflected by the prism 36,
The light is collected in the prism 38. The collected light is reflected by the prism 38, refracted by the lens system 40, and enters one end of the optical fiber 32. Light incident on one end of the optical fiber 32 irradiates the photodiode 23 of the operational amplifier 22. At this time, disturbance light is simultaneously generated in the photodiode 23.
To reach.

Then, a signal that has been photoelectrically converted by the photodiode 23 is input to the negative pole of the operational amplifier 22. This signal includes an electric signal of light emitted from the light emitting diode 16 and an electric signal of disturbance light. In addition, the positive pole of the operational amplifier 22 is connected to the light emitting diode 16 by “OF” from an integrating circuit composed of the amplifier 24 and the capacitor 44.
The offset voltage in the state of “F” is output. When the clock 12 turns off the light emitting diode 16 (see FIGS. 2A and 2B), at the same time, the first switch 18 is turned on and the second switch 20 is turned off. Therefore, the offset voltage includes an electric signal of disturbance light, electric noise generated by the operational amplifier 22, and the like (see FIG. 2B). Therefore, when the light emitting diode 16 is in the “OFF” state, the output of the operational amplifier 22 becomes 0 (see FIG. 2C).

The capacitor 44 stores an offset voltage. Therefore, when the light emitting diode 16 is turned on, the first switch 18 is turned off at the same time, but since the offset voltage is stored in the capacitor 44, the operational amplifier 22 is not affected by disturbance light or electric light. Then, an electrical signal of only the light of the light emitting diode 16 is output after removing noise and the like.

When the light emitting diode 16 is turned "ON", the second switch 20 is turned "ON" at the same time, and the capacitor 46 stores the voltage output from the operational amplifier 22. The capacitor 46 holds the accumulated voltage when the light emitting diode 16 is in the “OFF” state, that is, when the second switch 20 is in the “OFF” state. Thus, the signal output from the amplifier 26 has a curve shown in FIG.

The signal output from the amplifier 26 is read through the AD converter 70 of the light amount reading means 45A and is input to the comparator 74. The comparator 74 compares a signal output from the amplifier 26 with a signal of a threshold value described later output from the storage means 47, and outputs a signal to the one-shot 76 when each signal matches. One shot 76 outputs a C-SET signal based on a signal from comparator 74. Incidentally, the threshold value is stored in the storage means 47 in advance,
The threshold value is output to the comparator 74 via the DA converter 72.

Above the prisms 36 and 38, the blade 5
4 are provided, and the blade 54 is rotatably provided on the moving device 62. The moving device 62 moves in the Y-axis direction,
Since the blade 54 moves in the axial direction, the blade 54 also moves in the Y-axis direction and the Z-axis direction. The moving device 62 is provided with blade displacement reading means 45B, and the blade displacement reading means 45B reads the displacement of the blade 54 in the Z-axis direction. still,
The movement in the X-axis direction is performed by the base 31.

FIG. 3 shows a state in which the work 52 is being processed by the dicing apparatus 50, and FIG. 4 is an enlarged view of a main part of the blade position detecting apparatus 10. As shown in FIGS. 3 and 4, a pipe 56 for supplying water or air is provided near the inspection table 30, and a nozzle 56A is attached to a tip of the pipe 56. Then, during processing of the work 52 shown in FIG. 3, water is injected from the nozzle 63 and injected into a position where the blade 52 is processing the work 52. Thereby, the blade 54 is cooled, and the cutting property is further maintained.

On the other hand, since the blade 54 is rotating at a high speed during the processing, the chips of the work 52 generated by the processing become mist and drift near the processing position. In this case, water is ejected from the nozzle 56A through the pipe 56 of the blade position detecting device 10, and the opposing surfaces 36A of the prisms 36 and 38,
A water film is formed on 38A (see FIG. 5). Therefore, the mist does not adhere to the opposing surfaces 36A, 38A of the prisms 36, 38.

After processing a predetermined number of the workpieces 52, before moving the blades 54 in the X, Y, and Z-axis directions and arranging them between the prisms 36 and 38 of the blade position detecting device 10, air is supplied from the nozzle 56A for several seconds. Emit the prism 36, 38
Of each of the opposed surfaces 36A, 38A is air blown. Thus, the opposing surfaces 36A, 3A of the prisms 36, 38
8A is removed, so that the prism 3
Keep the opposing surfaces 36A, 38A of 6, 6 clean. Therefore, when the light guided from the prism 36 to the prism 38 passes through the facing surfaces 36A, 38A of the prisms 36, 38, the passing light is not refracted or scattered by the water droplet.

Further, a sequence may be created so that switching between the case where water is ejected from the nozzle 56A and the case where air is ejected is performed automatically, or the sequence may be manually operated. In FIGS. 1, 3, and 6, reference numeral 60 denotes a work table. The operation of the blade position detecting device according to the present invention configured as described above will be described in detail with reference to the flowchart shown in FIG.

During the processing of the work 52, water is injected from the nozzle 63 and injected into the processing position of the blade 54. Then, after processing a predetermined number of workpieces 52, before the blade 54 is disposed between the prisms 36 and 38 of the blade position detecting device 10, the injection of water from the nozzle 56A is stopped. next,
Air is ejected from the nozzle 56A for several seconds to form a prism 36,
Air blowing is performed on each of the facing surfaces 36A, 38A of the 38 (step 80). Next, the blade 54 is moved in the Y and Z axis directions and the base 31 is moved in the X axis direction, and the blade 54 is set at the inspection position of the opposing prisms 36 and 38 (step 82). Next, based on the clock 12, the light emitting diode 16 is alternately converted into the “ON” and “OFF” states, and the light amount data output from the amplifier 26 is read through the light amount reading unit 45A. Then, the process loops until the read light amount data continues for a predetermined time (step 84).

In this case, the operational amplifier 22 removes disturbance light, electric noise, and the like, and outputs an electric signal of only the light of the light emitting diode 16. After the loop is completed, it is checked whether the read light amount data is equal to or more than a predetermined value (that is, light amount data V1 when the prism 36 and the prism 38 are kept clean so that the blade 54 can be inspected) (step). 86). Next, the read light amount data is recorded in the memory as the initial light amount V0 (step 8).
8).

Next, the value of (V0 / 2) is set in the storage means 47A as a threshold value (step 9).
0). After the setting is completed, the blade 54 is lowered by one pulse (Z
(Moving downward in the direction of movement), the displacement of the blade 54 is read by the blade displacement reading means 45B (step 9).
2) Read the light amount data at that position (step 9)
4). After recording the read light amount data in the memory (step 96), 1 is added to the memory address (step 100).

Steps 92 to 106 are sequentially repeated until the blade 54 shields the entire sampling area 102 (ie, the area of the light beam reflected by the prism 36) 102 shown in FIG. During this repetition, the light amount data starts to decrease from the position where the blade 54 descends and reaches the sampling area 102. Next, when the light quantity data matches the threshold value, a C-SET signal is output from the one-shot 76 via the comparator 74 shown in FIG. The position of the blade 54 when the C-SET signal is output is defined as Zc, and this position is set as the reference position of the blade 54 to set the reference position Zc.
And the surface position Zt of the worktable 60 determined in advance.
Is stored (i.e., Zd = | Zc-Zt |).

The surface position Z of the work table 60 described above.
To obtain t, the blade 54 is attached to the worktable 60
, The blade 54 and the worktable 60 are energized, and the position of the blade 54 at this time is detected and set as the surface position Zt. Once the relative difference Zd is stored, if the reference position Zc is detected even when the blade 54 is replaced and the outer diameter of the blade is different, a new blade surface position Zt based on the stored relative difference Zd is obtained. (That is, Zt
= | Zc-Zd |) can be calculated. As a result, the uncut amount of the work 52 can be set as set.

When the blade 54 is further lowered, the entire area of the sampling area 102 is shielded from light, and the light amount data becomes zero. When the blade 54 is lowered in this manner, the outer diameter of the blade 54 is sufficiently large compared to the diameter of the sampling area 102, and the opposing surfaces of the prisms 36 and 38 are kept clean. Is located approximately in the middle between the position (Za) where the light intensity data starts to decrease and the position (Zb) where the light intensity data becomes 0 (see FIG. 8). That is, the relationship of (Za + Zb) / 2 ≒ Zc is established.

The relation between the value obtained by differentiating the (VZ) curve and dividing each differential value (ΔV / ΔZ) by (maximum differential value) and the position Z of the blade 54 is input in advance. (△ V / △) when the prisms 36 and 38 are in the clean state.
(Z-Z) It substantially matches the shape of the curve. Thus, it is determined that the prism 36 and the prism 38 are in a clean state (that is, a range where sufficient measurement accuracy can be obtained). Therefore, the position Zc of the blade 54 when the C-SET signal is output is set as the cutting position of the blade 54,
4 is moved onto the work 52, the blade 54 is set as a reference position, and the surface position Zt of the work table 60 is calculated from the previously obtained relative difference Zd. Then, the blade 54 is moved onto the work 52 and the blade 54 is moved to the work 5.
The workpiece 52 is cut by descending to a position where the set uncut amount is secured.

On the other hand, when dirt is unevenly adhered to the opposing surfaces of the prisms 36 and 38, Zc is not located in the middle between Za and Zb as shown in FIG.
The (△ V / △ Z-Z) curve is also input in advance (△ V / △ ZZ).
ΔZ-Z) Deviates from the shape of the curve. Therefore, in this case Z
The coordinate position of c is subjected to error processing as an unreliable value (steps 108 and 110).

When dirt adheres substantially uniformly to the opposing surfaces of the prisms 36 and 38, Zc is located at an intermediate position between Za and Zb as shown in FIG.
A ΔZ−Z curve is also input in advance (ΔV / ΔZ−
Z) It substantially matches the shape of the curve. Therefore, if the dirt is substantially uniform, even if the read initial light amount V0 is lower than a predetermined value (that is, light amount data when the prism 36 and the prism 38 are kept clean so that the blade 54 can be inspected). , To some extent (i.e., a range where sufficient repeatability can be obtained),
In this case, Zc is set as the reference position of the blade 54.
However, if the read initial light amount V0 becomes lower than the predetermined value by a certain degree or more, an error occurs in step 86.

As described above, according to the blade position detecting device of the present invention, the machine self-recognizes a state change due to dirt in the optical process of the optical sensor and an unstable state due to adhesion of water droplets, and makes the machine follow the changing state. If favorable conditions are not satisfied, retry or error processing can be performed. Therefore, disturbance light acting on the optical sensor and noise generated from the optical sensor can be removed, and the sensitivity of the optical sensor can be always kept constant. Accordingly, the blade wear amount can be accurately measured, and the blade tip can be positioned with high accuracy.

In the above embodiment, the threshold value is set to the initial light amount V0 / 2, but other values may be set to the threshold value. In the above-described embodiment, the space is made compact by using the prisms 36 and 38 in the blade position detecting device 10. However, the present invention is not limited to this, and the use of the prisms 36 and 38 may be omitted.

The blade position detecting device according to the present invention employs the size Z of the sampling area of the condensed light flux 102.
a and the difference between the actually measured size Zb of the light beam 102 | Z
By checking with a-Zb |, it can be confirmed whether or not the optical system is narrowed down to a spot diameter at which sufficient repetition accuracy can be obtained.

[0033]

As described above, according to the blade position detecting device of the present invention, the mist containing chips generated during the processing of a workpiece is prevented from adhering to the opposing surfaces of one and the other optical systems. Therefore, the opposing surfaces of the optical systems can be kept clean, and the amount of light guided from one opposing surface of the pair of optical systems to the light receiving means via the other opposing surface can be kept constant.

Therefore, when the amount of wear of the rotary blade changes due to conditions such as the type of work to be processed, the variation of the rotary blade, and the cutting depth, the rotary blade is in a non-contact state, that is, without contacting the processing table. The tip of the rotary blade can be positioned with high accuracy.

[Brief description of the drawings]

FIG. 1 is a schematic view of a blade position detecting device according to the present invention.

FIG. 2 is a diagram for explaining an operation state of each part of the blade position detecting device according to the present invention.

FIG. 3 is an enlarged view of a main part of the blade position detecting device according to the present invention.

FIG. 4 is a perspective view showing a state in which a blade of a dicing device is detected by a blade position detecting device according to the present invention.

FIG. 5 is a plan view illustrating a state in which water is emitted to the opposing surfaces of a pair of optical system prisms of the blade position detecting device according to the present invention.

FIG. 6 is a perspective view showing a state in which the blade of the dicing device is detected by the blade position detecting device according to the present invention.

FIG. 7 is a flowchart illustrating an operation state of the blade position detecting device according to the present invention.

FIG. 8 is a graph showing measurement results when the optical system of the blade position detecting device according to the present invention is free from dirt.

FIG. 9 is a graph showing measurement results when the optical system of the blade position detecting device according to the present invention is unevenly contaminated.

FIG. 10 is a graph showing measurement results when the optical system of the blade position detecting device according to the present invention is uniformly contaminated.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 ... Blade position detecting device 16 ... Light emitting diode (light projection means) 22, 24 ... Amplifier 23 ... Photodiode (light receiving means) 28, 32 ... Optical fiber 34, 40 ... Lens (a pair of optical systems) 36, 38 ... Prism (A pair of optical systems) 36A, 38A ... Opposing surfaces of prisms 44 ... Condenser 45A ... Light amount reading means 45B ... Blade displacement reading means 47 ... Storage means 54 ... Blade (rotary blade) 56A ... Nozzle 74 ... Comparator 76 ... One shot

Claims (1)

(57) [Claims] An optical system for converging light emitted from said light projecting means within said predetermined distance with one optical system, and diffusing after condensing; A pair of optical systems for condensing the collected light with the other optical system, light receiving means for receiving the light condensed by the other optical system and performing photoelectric conversion, and an optical axis between predetermined intervals of the optical system. A rotating blade that moves orthogonally to and blocks light condensed within the predetermined interval; a unit that reads the amount of light received by the light receiving unit and reads a displacement of the rotating blade; Means for outputting a signal when the signal coincides with the stored set value, and a blade position detecting device for positioning the tip of the rotary blade with respect to the workpiece based on the displacement position of the rotary blade when the signal is output. Wherein the blade position detecting device is a processing table for the rotary blade. A blade position detecting device provided apart from the optical system, further comprising: a washing unit that ejects water and air to one of the pair of optical systems and the other facing surface, wherein the washing unit ejects water at the time of processing and the optical system A water film was formed on one and the other opposing surfaces of the optical system to prevent the adhesion of chips generated during processing, and to eject air when detecting the position of the rotary blade and adhered to one and the other opposing surfaces of the optical system. A blade position detecting device for removing water droplets.