CN115070230A - Laser processing method and laser processing device for printed circuit board - Google Patents

Abstract

A laser processing method of a printed circuit board and a laser processing apparatus of a printed circuit board are provided, which effectively use a laser oscillator to optimize processing efficiency and quality. The laser processing method of the printed circuit board uses a laser oscillator for laser processing, wherein the laser oscillator oscillates laser by applying RF pulse. During the laser output period after the application of the RF pulse is completed, the laser oscillation is continued by applying the RF pulse again, and the laser beam of a desired time is extracted from the continuously oscillated laser beam to perform laser processing of the printed circuit board.

Description

Laser processing method and laser processing device for printed circuit board

Technical Field

The present invention relates to a method and an apparatus for processing a circuit board, and more particularly, to a method and an apparatus for processing a printed circuit board, which are suitable for forming a blind via (unperforated, hereinafter, simply referred to as a via or BH) in an insulating layer composed of an ABF material combined with a copper layer or an ABF with PET thereon in a process of packaging a substrate.

Background

In a conventional build-up printed circuit board, an insulating layer (hereinafter, simply referred to as "insulating layer") made of a resin containing glass fibers or fillers is integrated with a copper layer with the insulating layer interposed therebetween, and a hole of 40 to 120 μm for interlayer connection is formed by laser processing, and the hole is plated to connect the surface copper layer and the lower copper layer.

First, a structure of a conventional laser processing apparatus will be described.

Fig. 9 is a structural diagram of a conventional two-head laser processing apparatus.

A carbon dioxide laser oscillator 1 (hereinafter, referred to as a laser oscillator 1) outputs a pulsed linearly polarized laser light 2. The beam diameter adjusting device 3 disposed between the laser oscillator 1 and the beam splitter 4 is a device for adjusting the energy density of the laser light 2, and adjusts the energy density of the laser light 2 by changing the outer diameter of the laser light 2 output from the laser oscillator 1. That is, the energy of the laser beam 2 before and after the beam diameter adjusting device 3 does not change. Therefore, since the laser light 2 emitted from the beam diameter adjusting device 3 can be regarded as the laser light 2 output from the laser oscillator 1, the laser oscillator 1 and the beam diameter adjusting device 3 are collectively referred to as a laser output device 1A hereinafter. In addition, the beam diameter adjusting device 3 may not be used.

A beam splitter 4 is disposed between the beam diameter adjusting device 3 and the polarization conversion device 5A. The beam splitter 4 splits the laser 2 into two orthogonal directions, namely laser 2A and laser 2B. The laser light 2A is supplied to a first processing head not shown; the laser light 2B is supplied to a second processing head, not shown. Here, since the first machining head and the second machining head have the same structure, the objects (marks 5 to 12) having the same structure are distinguished by adding a letter A, B, that is, marks 5A, 6A, 7A, 8A, 9A, 10Aa, 10Ab, 11A, and 12A are structures applied to the first machining head, marks 5B, 6B, 7B, 8B, 9B, 10Ba, 10Bb, 11B, and 12B are structures applied to the second machining head, and the objects corresponding to the marks have the same structure. Hereinafter, only the case of the first processing head will be described.

The polarization conversion device 5A converts the linearly polarized laser light 2A into circularly polarized laser light 6A. The polarization conversion device 5A includes a reflected light blocking mechanism (details are omitted) that blocks the laser beam 6A reflected by the processing portion during processing and has a function of preventing damage to the laser oscillator 1 caused by the laser beam 6A reflected by the processing portion. The flat plate 7A disposed between the polarization conversion device 5A and the galvanometer mirror 10Aa is made of a material (e.g., copper) that does not allow the laser beam 6A to pass therethrough, and a plurality of apertures 8A (windows, in this case, circular through holes) are selectively formed at predetermined positions. The plate 7A is driven by a driving device, not shown, and positions the axis of the selected aperture 8A coaxially with the axis of the laser 6A. The galvanometer device 9A is configured by a pair of galvanometer mirrors 10Aa and 10Ab, and is rotatable around a rotation axis as indicated by arrows in the drawing, and can position a reflection surface at an arbitrary angle. The f θ lens (condenser lens) 11A is provided in a first processing head (not shown). The galvanometer mirrors 10Aa, 10Ab, and the f θ lens 11A constitute an optical axis positioning device that positions the optical axis of the laser light 6A at a desired position 12A in the printed circuit board, and a scanning area (i.e., a processing area) 12A defined by the rotation angles of the galvanometer mirrors 10Aa, 10Ab and the diameter of the f θ lens 11A is about 50mm × 50mm in size. A printed circuit board 13, which is a workpiece and is composed of a copper layer and an insulating layer, is fixed to an X-Y table 14. The first processing head and the second processing head may process printed circuit boards 13 of the same pattern, or may process printed circuit boards 13 of different patterns. The control device 20 controls the laser oscillator 1, the beam diameter adjusting device 3, the driving devices of the flat plates 7A and 7B, the galvanometer mirrors 10Aa, 10Ab, 10Ba, and 10Bb, and the X-Y stage 14 (the first processing head and the second processing head may correspond to one X-Y stage, respectively) in accordance with the inputted control program.

Then, when the hole processing is performed, the X-Y stage 14 is moved so that the designated processing regions 12A and 12B face the f θ lenses 11A and 11B, respectively, and then, after the holes are processed by performing one beam irradiation (i.e., one pulse irradiation) on all the copper layers in the processing regions 12A and 12B, the insulating layer is processed by one or more pulse irradiation, thereby completing the holes in the processing regions 12A and 12B.

Referring to fig. 9 and 10, fig. 10 is a diagram showing the setting time of the galvanometer mirror and the laser irradiation time, and the horizontal axis in the diagram is time.

In fig. 9, the time during which the positioning time becomes longer in a certain processing position in the galvanometer mirror 10Aa and the galvanometer mirror 10Ab of the first processing head is referred to as a galvano time GA of the first processing head (denoted by GA1 and GA2 in fig. 10). The time during which the positioning time becomes longer in a certain processing position of the galvanometer mirror 10Ba and the galvanometer mirror 10Bb of the second processing head is referred to as a stream sensing time GB (indicated as GB1 and GB2 in fig. 10) of the second processing head. L1 represents the laser irradiation time of one shot.

As shown in fig. 10 (a), when the printed circuit boards 13 processed by the first processing head and the second processing head (or the first processing head and the second processing head correspond to one printed circuit board each) have the same processing content and the same flow detection time, the laser light 2 can be supplied to both processing heads at the same time. However, as shown in (a) and (b) of fig. 10, when the processing contents of the first processing head and the second processing head are different or the current detection time GA and GB are different, it is necessary to match a long current detection time in order to supply the laser light 2 to the first processing head and the second processing head at the same time. That is, when GA1< GB1, GA2> GB2, a wait time of (GB1-GA1) may occur at the first processing head, and a wait time of (GA2-GB2) may occur at the second processing head. Therefore, the overall processing efficiency is lowered.

The problems to be solved by the invention are:

in recent years, with the progress of thinning of substrates, a hole processing is performed on an insulating layer having no copper layer on the surface thereof to form a substrate for packaging. In this case, when the diameter of the hole to be processed is 60 μm or less, the output required for processing is about 20W. Next, an actual processing example will be described.

FIG. 11 illustrates an example of laser irradiation with the vertical axis representing the oscillated laser output; the horizontal axis is time. The upper stage shows switching of an RF pulse (hereinafter, simply referred to as RF) for exciting the oscillation medium of the laser oscillator 1.

For example, when a hole of 60 μm is to be formed in an insulating layer, the laser beam is irradiated for 20 μ s at an RF time according to an output curve a1 in which the output is adjusted to 20W, and then the irradiation with the laser beam is continued once again under the same conditions as the pulse period. That is, the processing time in this case is such that the pulse period of the first pulse irradiated after the galvanometer mirror is positioned is 100 μ s, and the sum of the laser light in the second pulse period 100 μ s after the first pulse and the subsequent 80 μ s (including the time 60 μ s from the laser light after RF stop to extinction) becomes 180 μ s. Here, unlike the case of processing a copper layer having a high processing threshold or processing an insulating layer having a thickness of more than 60 μm after the copper layer is processed, since the insulating layer having a low processing threshold and a thickness of only about 30 μm is processed, the amount of heat accumulated in the processing portion is small, and thus 2 times of continuous irradiation with laser light can be performed.

Further, as shown in fig. 11, the output curve a1 is a curve having the 1 st peak output immediately after RF application (the duration of the output is a short time). This 1 st peak output is about 1/2 of the 2 nd peak output in the standard RF application time (in this case 20 μ s). After the RF application is stopped, the energy accumulated in the oscillation medium in the laser oscillator is converted into laser light. In the illustrated state, the duration of the laser beam emitted by the oscillating medium in the laser oscillator is about 60 μ s.

However, the above processing causes the following problems.

(1) The diameter of the inlet of the processed hole may be increased by 2 to 3 μm due to the influence of the 1 st peak output.

(2) The diameter of the processed hole is not uniform due to the output variation of the laser.

Here, the output fluctuation of the laser light will be described by using an illustration.

FIG. 12 is a graph showing the output variation of the 2 nd peak output, wherein (a) in FIG. 12 shows the condition of the pulse frequency of 1-5 KHz, and (b) in FIG. 12 shows the condition of the pulse frequency of 1-10 KHz. For example, when the intervals of the holes to be processed are substantially the same, the 2 nd peak output hardly varies. However, the intervals of the holes to be processed are determined by the intervals of the mounted products mounted on the printed circuit board or the positions of the holes connected to the copper layer of the lower layer, and thus are not uniform. Thus, the gain (gain) accumulated in the excitation medium is changed due to the change of the laser excitation interval (i.e., duty cycle), and the output at the frequency of 1 to 5KHz is changed by about ± 3% as shown in (a) of fig. 12, and the output at the frequency of 1 to 10KHz is changed by about ± 5% as shown in (b) of fig. 12. Therefore, the diameters of the holes to be processed may vary.

(3) In the RF stop period after the RF application time of 20 μ s, the energy accumulated in the laser medium is irradiated to the processing portion during a period from the RF stop to about 60 μ s, and the temperature of the processing portion is increased. Even if the temperature rise value of the processed portion is lower than the processing threshold value of the insulating layer, if the temperature rise value is continued for 10 μ s or more, the insulating layer at the hole bottom and the hole wall surface is easily carbonized, and the processing quality of the hole is deteriorated.

As described above, it is desired to make the processed pore diameter uniform and prevent the quality of the insulating layer from being degraded.

The invention aims to provide a laser processing method of a printed circuit board and a laser processing device of the printed circuit board, which have excellent processing efficiency and excellent quality by effectively activating a laser oscillator.

Disclosure of Invention

An object of the present invention is to provide a laser processing method of a printed circuit board and a laser processing apparatus of a printed circuit board, which effectively use a laser oscillator to optimize processing efficiency and optimize quality.

To solve the above problem, a first means of the present invention is:

a laser processing method of a printed circuit board, which performs laser processing using a carbon dioxide laser oscillator that oscillates laser light by applying an RF pulse, characterized in that:

during the laser output period after the application of the RF pulse is completed, the laser oscillation is continued by applying the RF pulse again, and the laser beam of a desired time is extracted from the continuously oscillated laser beam to perform laser processing of the printed circuit board.

A second means of the present invention is the laser processing method for a printed circuit board described above, wherein a saw-tooth pulse is generated by controlling an application time of the RF pulse and a stop time of the RF pulse, and processing is performed by the generated saw-tooth pulse.

A third means of the present invention is the laser processing method of a printed circuit board described above, wherein at least one of a sum of an application time of the RF pulse and a stop time of the RF pulse, or a ratio of the application time of the RF pulse and the stop time of the RF pulse is controlled.

The fourth means of the present invention is:

a laser processing apparatus for a printed circuit board performs laser processing using a carbon dioxide laser oscillator that oscillates laser light by applying an RF pulse, the laser processing apparatus including a control device.

The control device performs the following control: during the laser output period after the application of the RF pulse is completed, the laser oscillation is continued by applying the RF pulse again, and the laser beam of a desired time is extracted from the continuously oscillated laser beam and supplied to the processing unit.

A fifth aspect of the present invention is the laser processing apparatus for a printed wiring board, further comprising a plurality of processing heads, and a mechanism for distributing laser light from the carbon dioxide laser oscillator to each of the processing heads, wherein when positioning of any one of the processing heads is completed, the control device supplies laser light to the processing head without considering a positioning state of the other processing head.

The invention has the beneficial effects that: by the method and the device, the hole diameter of the processed hole and the quality of the wall surface of the hole can be improved. Moreover, the processing efficiency can be improved. Further, since the laser beam is continuously oscillated stably, when the laser beam is supplied to a plurality of machining heads, the necessary laser beam can be supplied to any one of the machining heads regardless of the other machining head, and thus the machining efficiency of the laser machining device can be improved. Therefore, the hole diameter of the processed hole is uniform, the quality of the wall surface of the hole is improved, and the processing efficiency is improved. Further, since stable laser light can be continuously oscillated and necessary laser light can be supplied without affecting other machining heads when stress light is supplied to a plurality of machining heads, the machining efficiency of the laser machining apparatus can be improved.

Drawings

FIG. 1 is a drawing illustrating the essential elements of a sawtooth wave of the present invention;

FIG. 2 is an example of an output waveform;

FIG. 3 is a diagram illustrating a generating step of a sawtooth wave according to the present invention;

FIG. 4 is a graph illustrating energy at the time of opening a hole;

fig. 5 (a), (b) are diagrams illustrating the spatial distribution of energy of the output;

FIG. 6 is a structural view of an embodiment (two-head type laser processing apparatus) of the laser processing apparatus of the printed wiring board of the present invention;

fig. 7 (a) and (b) are diagrams showing the setting time of the galvanometer mirror and the laser irradiation time;

fig. 8 (a), (b), (c), and (d) are diagrams illustrating rectangular pulses when processing an insulating layer containing a filler (filler);

fig. 9 is a structural view of a conventional two-head laser processing apparatus;

fig. 10 (a) and (b) are diagrams showing the setting time of the galvanometer mirror and the laser irradiation time;

FIG. 11 is a drawing for explaining an example of laser irradiation; and

fig. 12 (a) and (b) are diagrams showing output changes in the 2 nd peak output.

Detailed Description

Fig. 1 is a diagram illustrating constituent elements of a sawtooth wave according to the present invention.

The output curve C (fundamental wave pulse waveform) in fig. 1 is an output curve having a nominal duty cycle of 60% (i.e., RF application time/pulse cycle, hereinafter, the nominal duty cycle is represented by "Dty"), a frequency of 10KHz, and a maximum output of 250W.

First, the output curve C is explained. When the RF is activated at time T0, the laser beam is explosively output at time T1, and after the 1 st peak output is reached at time Tj, the laser beam is attenuated to time Td, and then increased again, and after the RF application time 60 μ s has elapsed from time T0, the laser beam reaches the 2 nd peak output 250W at time T2. Here, the output shown by the solid line in the RF application time is the sum of the output of the N2 gas excited by RF and transferred to the CO2 gas through the excitation medium (the output shown by the broken line in the vicinity of 60 μ s in this case) and the output of the CO2 gas directly excited by RF (that is, the output shown by the broken line and the output shown by the solid line). Here, when the RF application is stopped, the output of the CO2 gas directly subjected to the RF excitation becomes 0, and the energy accumulated in the N2 gas as the excitation medium in the laser oscillator is output as laser light after time T2. Further, the energy accumulated in the excitation medium is output for about 60 μ s after time T2.

The present inventors first confirmed the following conditions by experiments and simulations:

1. element 1

For example, when the oscillation is performed with the Dty 60% (pulse width 60 μ s) and the RF of 10KHz applied, the oscillated laser output slowly increases and becomes maximum at the pulse width 60 μ s as shown in the output curve C. After the RF application is stopped, the output is attenuated. When the laser oscillator oscillates within the range of Dty (RF application time/RF cycle), the output rises along the same output curve (output curve C in fig. 1) even if the RF application time is changed. That is, when the oscillation is performed at a Dty of 40% (pulse width 40 μ s) and 10KHz, the output rises along the output curve C, and the pulse width is maximized at 40 μ s, and the attenuation is performed in the same manner as described above. When the oscillation is performed under the conditions of Dty 20% (pulse width 20 μ s) and 10KHz, the oscillation rises along the output curve C, reaches the maximum value at a pulse width of 20 μ s, and attenuates in the same manner as in the above-described case. In fig. 1, a rising portion of the output is indicated at P1, and a switching portion of the output is indicated at P2.

2. Element 2

When the carbon dioxide laser is used, if the application of RF is started at time T0, the laser starts to oscillate due to the excitation output accumulated in the excitation medium at time T1, the output rapidly increases, and after the 1 st peak output Wj is reached at time Tj, the output once attenuates (time Td) and thereafter increases again, and becomes the 2 nd peak output when the application of RF is stopped.

At this time, the time Td is 0.4 to 0.5 μ s from the time T0, and even if Dty and the pulse width change, it can be seen that the output response (output change per unit time) Ws at the time Td is substantially constant.

3. Element 3

After the RF application is stopped, the laser output is switched from 250W to an output based on the residual energy accumulated in the laser medium. The switching time is 0.4-0.5 μ s, the output decreases, then increases a little, and then decays. Hereinafter, an output response when the output falls at the time of output switching is referred to as an output response Wc, and an output response when the output slightly rises is referred to as an output response Wd. This output response Wc is a value that is almost constant even if Dty and the pulse width are changed. The output response Wc and the output response Ws have different output directions, but the magnitudes of the output components are almost the same.

4. Element 4

FIG. 2 shows an example of an output waveform, i.e., an output waveform when the energy accumulated in the waveform excitation medium starts applying RF again at the end of the pulse period. As shown in FIG. 2, if there is a residual output accumulated in the excitation medium at the start of RF application (i.e., Δ z shown by the oblique line in FIG. 2, and the output at the start of RF application is Δ w. here, Δ w >0), the output excited at the start of excitation of the next (2 nd) pulse period overlaps with the residual output as shown by the oblique line, and the output of the output response Ws increases after 0.4 to 0.5 μ s. And the 1 st peak output Wj does not occur at this time. Although the description is given of the case between the 1 st pulse period and the 2 nd pulse period, when the pulse period is repeated for the nth pulse period, the output intensity of the output response Ws gradually increases, and the 2 nd peak output gradually decreases relatively, but becomes stable after 1 second.

In the situation of fig. 11, the output curve of the 2 nd pulse period is the same as the output curve of the 1 st pulse period because the surplus energy accumulated in the excitation medium before the 2 nd pulse period starts becomes 0.

Referring to fig. 1, the value of the output integral value continuously oscillating at a Dty of 60% (i.e., the total energy output in the pulse period of 0-100 μ s) divided by the pulse period of 100 μ s in the output curve C is referred to as the average output Wav. In addition, when the Dty is fixed and the pulse period is in the range of 20 to 200KHz, the average output Wav shown in FIG. 1 is almost constant. On the other hand, when the pulse period is fixed, the average output Wav increases in proportion to the Dty. The output response Wr (see fig. 3) and the output response Wf (see fig. 3) in the tangential direction of the output curve C are values specific to the preset average output.

Fig. 1 shows 100KHz and 200KHz sawtooth waveforms, where P3 in fig. 1 is an output rising portion, P4 is an output switching portion, and P5 is a next output rising portion. The details thereof will be described with reference to fig. 3.

FIG. 3 is a drawing for explaining a generating step of a sawtooth wave according to the present invention.

When Dty is set to trf1/tm (tm is a pulse period and is constituted by the RF application time trf1 and the RF application stop time trf 0) in the output curve C (see fig. 1), the average output Wav is determined. Even if the pulse period tm is shortened and the ratio of the RF application time trf1 to the RF application stop time trf0 is the same, the Dty is the same and the average output Wav does not change. Here, the sawtooth wave of the present invention is generated based on the fundamental wave pulse waveform (the output curve C) and the elements 1 to 4.

The generation of the saw-tooth wave is described below. The sawtooth pulse of the present invention is generated in the following steps based on the fundamental wave pulse waveform (the above-described output curve C).

The method comprises the following steps: with the vertical axis as the output axis, Dty, pulse period tm, average output Wav, and upper limit output Wp are set. Here, the upper limit output Wp is an output (J/s) obtained enough for the irradiation pulse to obtain the target aperture, and is set to a value corresponding to the material threshold. The lower limit output Wv is an output greater than the processing threshold Wm of the insulating layer.

Step two: the horizontal axis represents a time axis, and a point on the lower limit output Wv at time t0 represents Q1, and a point on the lower limit output Wv at time t2 representing a pulse period represents Q6. Then, the output response Ws with the point Q1 as the starting point is denoted, and the end point of the output response Ws is the point Q2.

Step three: the point Q2 is connected to the point Q3 on the upper limit output Wp at time t1 with the output response Wr. The time t1 is the end of trf1 (the start point of trf 0).

Step four: the point Q3 is taken as the starting point and the output response Wc is marked, and the end point of the output response Wc is taken as the point Q4.

Step five: a small output response Wd is indicated on the extended line connecting the point Q1 and the point Q4, and the end point is taken as the point Q5.

Step six: the point Q5 and the point Q6 are connected by an output line Wf. That is, the point Q5 is the intersection of the extension line connecting the points Q1 and Q4 and the output response Wf having the point Q6 as the starting point.

Through the above steps, the sawtooth wave applied to the lower limit output Wv is completed.

Hereinafter, the sawtooth pulse superimposed on the lower limit output Wv using the polygon formed in the above step is referred to as a "sawtooth pulse".

As will be described later (… refers to "rectangular pulse with saw teeth n" when the pulse period tm is changed at a fixed Dty), the output of the laser oscillator becomes a steady state after more than 1 second, and the range of change between the average output Wav and the upper limit output Wp is about ± 1% as long as the laser oscillator is kept in an operating state.

FIG. 1 shows a 100KHz sawtooth pulse and a 200KHz sawtooth pulse generated by the above steps.

Further, as shown in fig. 3, when the point Q1 and the point Q4 are connected by a broken line, the output surrounded by the quadrangle Q1Q2Q3Q4 corresponds to the output of the direct RF excitation of the CO2 gas as described above (fig. 1 shows that the energy accumulated in the excitation medium in fig. …, which explains the constituent elements of the saw-tooth wave of the present invention, is about 60 μ s after the time T2).

When the pulse period tm is changed at a fixed Dty, the output responses Ws and Wc do not change. On the other hand, although the output responses Wr, Wd, Wf vary in response to the pulse period tm, the average output does not vary.

When Dty is changed at a fixed pulse period tm, the output responses Ws and Wc hardly change, and the changes in the output responses Wr, Wd and Wf are small. The average output varies according to the Dty, but is set to be an intrinsic value.

Therefore, by setting the pulse period tm, the RF application time trf1, and the RF application stop time trf0 in the Dty range, the waveform and output of the sawtooth pulse can be controlled.

In actual machining, the waveform generation step is performed continuously to generate n sawtooth pulses (n is an integer of 1 or more, and hereinafter referred to as "rectangular pulses of sawtooth n").

Fig. 4 is a graph illustrating energy at the time of drilling, with output on the vertical axis and time on the horizontal axis.

Here, assuming that the holes to be processed are the same as those described in fig. 11, the conventional technique performs processing with 2 pulses having a pulse frequency of 10KHz, and has a laser irradiation time of 20 μ s for 2 times, which is 180 μ s, which is the sum of the pulse period of the 1 st pulse of 100 μ s and the RF application time of the 2 nd pulse of 20 μ s and the non-excitation time of 60 μ s. On the other hand, in the present invention, as shown in fig. 4, since the pulse energy equivalent to that of the conventional art can be obtained by processing the rectangular pulses of the saw-tooth 2 having the 2 pulse frequency of 100KHZ without being affected by the pulse period, the interval of the 2 rectangular pulses can be arbitrarily set, and even if the interval of the 2 saw-tooth rectangular pulses is set to 60 μ s (the conventional pulse interval is 20 μ s), the processing time is 100 μ s. Therefore, according to the present invention, the processing time can be shortened by 80 μ s compared with the conventional one.

Next, the present invention and the prior art will be described with reference to the shape of the machined hole.

Fig. 5 is a diagram illustrating a spatial distribution of output energy, where fig. 5 (a) shows the case of the present application and fig. 5 (b) shows the case of the related art. The vertical axis represents the normalized energy level and the processing depth, and the horizontal axis represents the radial direction of the hole. Wherein Ds represents a spot diameter of the processing portion, DR represents a target aperture, DR1, DR 'represent apertures smaller than DR, and DB, DB' represent hole bottom diameters. Lv0 represents the position of energy level 0, Lv1 represents the surface position of the insulating layer, Lv2 represents the bottom surface position of the insulating layer, and k represents the processing threshold of the insulating layer. Ep represents the energy distribution when Wp is output, ev represents the energy distribution when Wv is output, eav represents the energy distribution when Wav is output on average, and 1e represents the energy distribution of the first pulse; and 2e represents the energy distribution of the second pulse. The outputs Wp and Wv and the average output Wav are shown in fig. 3.

Referring to fig. 3 and 5, the processing steps of the present invention will be described below with respect to the radial direction of the processed hole. The energy distribution when the machining diameter increases at the RF application time trf1 is referred to as a diameter increasing energy distribution, and the energy distribution when the machining diameter decreases at the RF application stop time trf0 is referred to as a diameter decreasing energy distribution. The processing is started with an output response Ws superimposed on the output Wv at the same time as the RF application and increased by about 0.4 μ s, and then the processing is performed with an energy distribution increased according to the processing diameter of the output response Wr, and after a time trf1, the entrance diameter of the target hole is formed. When the RF application is stopped, the machining is performed based on the output response Wc of the machining diameter reduction energy distribution, the output response Wd of the machining diameter fine increase, and the output response Wf of the machining diameter reduction energy distribution.

In the above processing step, the processing is performed by alternating the spot diameters DR and DB and DR 'and DB'. The output rise at the start of machining is steeper than that of the conventional pulse and the irradiation time is short. Further, since the energy distribution diameter after the RF application is stopped is reduced and is away from the hole entrance and the hole side wall, the thermal conduction between the hole entrance and the hole side wall surface during the processing of the insulating layer is reduced, so that the thermal influence of the insulating layer on the hole wall surface is reduced, and the hole quality is improved. Further, since the zigzag pulse wave processing is continuously performed, laser irradiation which is not related to the processing in the conventional art is not performed. Therefore, the mass of the hole entrance and the aperture wall is reduced. In addition, the diameter of the entrance to the hole does not expand because it is not affected by the 1 st peak output.

Fig. 6 is a configuration diagram of an embodiment of a laser processing apparatus for a printed circuit board (two-head laser processing apparatus) according to the present invention, and the same reference numerals are used for the same objects or objects having the same functions as those in fig. 9, and detailed description thereof is omitted.

By setting the application time and stop time of the high frequency RF for driving the laser oscillation, one laser oscillator 1 outputs one laser beam 2 having a continuous linearly polarized saw-tooth shape with a frequency of 50KHz or more. A beam diameter adjusting device 3 disposed between the laser oscillator 1 and a beam splitter 4 is a device for adjusting the energy density of the laser light 2, and adjusts the energy density of the laser light 2 by changing the outer diameter of the laser light 2 output from the laser oscillator 1. That is, the energy of the laser beam 2 before and after the beam diameter adjusting device 3 does not change. Therefore, since the laser light 2 emitted from the beam diameter adjusting device 3 can be regarded as the laser light 2 output from the laser oscillator 1, the laser oscillator 1 and the beam diameter adjusting device 3 are referred to as a single laser output device 1A. In some cases, the beam diameter adjusting device 3 is not used.

An AOM50A driven by a driving member 61A (a driving member 61B drives an AOM50B) is disposed between the beam splitter 4 and a polarization conversion device 5A. The AOM50A divides one laser 2A into one 1 st light laser 2A1K and one 0 th light laser 2A0 (the AOM50B divides one laser 2B into one 1 st light laser 2B1K and one 0 th light laser 2B0), and adjusts the output of the laser 2A1K for processing by changing the proportion (opening degree) of the allocation. The laser light 2a0 not used in the processing is not diffused to the periphery as much as possible and is discarded in a buffer not shown.

The laser machining device is configured to be able to position a second machining head in the X direction with respect to a fixed first machining head by a second machining head moving device, not shown. The laser machining device is also configured to be able to extend the position of the second machining head by a maximum distance s relative to the first machining head. One mirror 31 and one mirror 34 are fixed at predetermined positions, and the two mirrors 32, 33 are supported by a mirror moving device, not shown, and can be freely positioned in the X direction. The mirrors 31 to 34 are arranged such that the axis of one aperture 8B coincides with the center of one galvanometer mirror 10Ba regardless of the position of the mirrors 32 and 33 in the X direction. A control device 20 for controlling the laser oscillator 1, the beam path adjusting device 3, the drives 61A, 61B of the AOM, the drives of the plurality of flat plates 7A, 7B, the galvanometer mirrors 10Aa, 10Ab, 10Ba, 10Bb, and an X-Y stage 14 (optionally, the first processing head and the second processing head may correspond to one X-Y stage, respectively); not shown are a second machining head movement device and a mirror movement device.

The processing steps are explained below. The description will be given of the case of the first machining head since the machining contents are different for each machining head and the operation is substantially the same.

When the start of machining is instructed, the control device 20 drives the moving device of the second machining head to move the second machining head to the specified position. Next, the X-Y stage 14 is controlled to position the first processing head at a processing position, and the galvanometer mirrors 10Aa, 10Ab are positioned at the starting processing position and stand by. In addition, a movement device, not shown, of the second machining head is actuated to move the second machining head a distance s relative to the first machining head. Next, the mirror moving device, not shown, is operated to move the position of the mirror 32 by a distance s/2 in the moving direction of the second processing head. In this manner, since the distance between the diaphragm 8B and the galvanometer mirror 10Ba becomes constant, the size of the image of the diaphragm 8B can be kept constant regardless of the position of the second processing head.

Then, the laser oscillator 1 is first activated, and after a predetermined waiting time has elapsed, the machining program is started to start machining. The reason why the waiting time is set is that the output of the laser oscillator 1 is unstable until thermal equilibrium is reached, and the time is about 1 to 2 seconds.

After the elapse of the waiting time, the control device 20 outputs the laser light 2 (hereinafter, simply referred to as laser light 2) shaped in a zigzag form from the laser oscillator 1 in accordance with a machining program input in advance. The laser beam 2 is changed in diameter by the beam diameter adjusting device 3, and is split into the laser beam 2A by the beam splitter 4, and is incident on the AOM 50A. The AOM50A discards the laser beam 2A in the buffer until it receives an operation command from the control device 20. After receiving a positioning completion signal of the galvanometer mirror that completes positioning later in the galvanometer mirrors 10Aa and 10Ab, the controller 20 actuates the AOM50A by the driving member 61A to output the laser beam 2A as a square pulse 2A1K (i.e., the above 1-time laser beam 2A1K), and the square pulse 2A1K is composed of n saw-tooth pulses attenuated to a predetermined output. The rectangular pulse 2A1K is positioned by the galvanometer mirrors 10Aa, 10Ab and is incident on a specified position of the printed circuit board 13 to form a hole in the printed circuit board 13. Thereafter, the above hole drilling operation is repeated as in the conventional case until the specified machining is completed. In the above processing, the laser oscillator 1 continuously outputs the laser light 2 from the start to the end of the processing by turning on and off the RF at predetermined cycles and pulse cycles.

Fig. 7 is a diagram showing the setting time of the galvanometer mirror and the laser irradiation time in the present invention, wherein (a) in fig. 7 shows a case where the first processing head is used, and (b) in fig. 7 shows a case where the second processing head is used, and the horizontal axis in fig. 7 is time. As shown in fig. 7, the laser beams 2A and 2B composed of saw-tooth pulses are output in a normal state during machining, and in the case of the first machining head, after the current detection times GA1 and GA2 are completed, the rectangular pulse 2A1K composed of n saw-tooth pulses necessary for the machining is irradiated to the machining portion through the AOM50A to form a hole. At this point, the first machining head may continue to operate without regard to the current sensing time GB1, GB2 of the second machining head. Likewise, the second machining head may continue to operate without regard to the flow detection times GA1, GA2 of the first machining head. Therefore, as each processing head does not need waiting time, compared with the prior art, the processing efficiency can be improved by 20-30%. After the positioning of the galvanometer mirror is completed based on a control clock signal of an apparatus not shown in the drawing, the AOM50 is controlled so that the rectangular pulse of the saw tooth n supplied to the processing unit is not chipped or missed by outputting the RF application start timing in accordance with the first saw tooth pulse.

Next, the structure of the saw blade of the present invention will be described in detail.

Fig. 8 is an explanatory diagram illustrating a rectangular pulse when processing an insulating layer containing a filler (reinforcing material), where the horizontal axis represents time and t is a time based on t 0.

In FIG. 8, the upper section shows the operation of the AOM, where A is 100% of the opening of the AOM and mA is m% of the opening of the AOM. The middle section shows on or off of the RF, tm is the pulse period, trf1 is the on period of the RF, trf0 is the off period of the RF. The lower panel shows an output of Wpf, which is an upper limit output for filler (filler) processing of the insulating layer, and Wpr, which is an upper limit output for resin processing of the insulating layer.

In fig. 8 (a), the laser beam path is enlarged during RF on period to process mainly the filler, and the laser beam path is reduced during RF off period to remove the gas or machining debris generated by the processing from the processing portion quickly, thereby improving the processing quality of the hole wall surface and the hole bottom. In addition, in fig. 8 (b) and 8 (d), as in fig. 8 (a), gas or machining chips generated by machining can be quickly removed from the machining portion, and therefore, the machining quality of the hole wall surface and the hole bottom can be improved.

Fig. 8 (c) shows a variation of the portion surrounded by the broken line in fig. 8 (a), and fig. 8 (a) shows an example of waveform control during the output rise in the RF start period (from time td1 in the figure, AOM is started only at time ta). Therefore, the average output Wavh of the output ascending part is used for processing, so that the hole inlet step and the hole wall are uniform, and the hole quality is improved. Furthermore, tg is the time of AOM shutdown, tp is the time point of the initial start of AOM.

For example, when the hole is inclined in the depth direction, (d) in fig. 8 is used.

Next, an example of the processing will be described.

The results of processing an insulating layer (ABF material manufactured by ajinomoto corporation, thickness about 30 μm) containing a sealing filler with a pore diameter of 60 μm in the same manner using the rectangular pulses of saw teeth n in (a), (b) and (c) in fig. 8 or the conventional pulses in fig. 11 are as follows. In addition, fig. 8 is a diagram showing the shape of the sawtooth pulse, not the rectangular pulse showing the sawtooth n for machining.

In the case of fig. 8 (a), the saw blade 3 was machined under the conditions of 2 rectangular pulses with a frequency of 100KHz, a Dty of 60% (trf1 ═ 6 μ s, trf0 ═ 4 μ s), and an AOM opening of 100%, and the hole diameter-to-bore diameter was about 62 μm and the hole (bottom diameter/diameter-to-bore diameter) ratio was about 80%.

In the case of fig. 8 (b), the saw blade 3 was machined under the conditions of 2 rectangular pulses with a frequency of 100KHz, a Dty of 60% (trf1 ═ 6 μ s, trf0 ═ 4 μ s), and an AOM opening of 0%, and the hole diameter-to-bore diameter was about 60 μm and the hole (bottom diameter/diameter-to-bore diameter) ratio was about 80%.

In the case of fig. 8 (c), 2 rectangular pulses of saw blade 3 having a frequency of 100KHz, Dty 60%, td1 ═ 6 μ s, were processed under the condition of 0% AOM opening, or ta ═ 4 μ s, was processed under the condition of 100% AOM opening, and the hole entrance diameter was about 60 μm and the hole (bottom diameter/entrance diameter) ratio was about 81%.

In the cases (a), (b), and (c) in fig. 8, resin carbonization was hardly observed on the surface of the copper layer at the bottom of the hole.

In the case of the conventional pulse shown in FIG. 11, the hole entrance diameter was about 65 μm and the hole (bottom diameter/entrance diameter) ratio was about 77% by processing under the conditions of a pulse width of 20 μ s, a pulse frequency of 10kHz and a pulse number of 4. In addition, the copper surface at the bottom of the hole was found to have resin carbonized. In addition, when the pulse has the 1 st peak output Wj higher than the 2 nd peak output Wp, the output during processing is increased, and thus the ring-shaped damage is generated around the entrance of the hole due to the high refracted light of the 1 st peak output Wj.

Referring to fig. 3, the values of the optimal output responses Ws, Wc, Wr, Wd, and Wf vary depending on the material of the workpiece. Therefore, by setting the values of the output responses Ws, Wc, Wr, Wd, and Wf to the optimum values according to the material of the workpiece, the processing quality and the processing speed can be improved.

Incidentally, in actual machining, the values of the output levels Wp and Wv and the values of the output responses Ws, Wc, Wr, Wd and Wf corresponding to the respective workpieces can be known in advance. When the rated duty cycle and the pulse cycle are determined, the maximum output of the laser oscillator can be known in advance. Further, an appropriate aperture diameter corresponding to the processing aperture is also known.

Here, if the first material is processed, for example, the numerical values of the energy levels Wp and Wv are temporarily set with reference to the existing data, and the numerical value is determined by comparing the experimentally processed aperture with the intended aperture and increasing or decreasing the output energy level test Wp. Then, the sawtooth pulse n is used to process, and the value of n is determined according to the depth of the processed hole. In this case, since the quality is deteriorated by heat of the insulating layer when n is increased, if n is large, n is divided into rectangular pulses of a plurality of saw-teeth m, and a time for cooling the processing portion is provided between the rectangular pulses.

In the two-head laser beam machine shown in fig. 6, since the average output of the laser oscillator 1 is 125W when the output is 250W, the output is divided by the beam splitter 4 and 62.5W is supplied to each machining head. Therefore, as described above (the following description explains that the processing example … causes annular damage around the hole entrance), when Wpf W is 20W, both the first processing head and the second processing head can be used for processing. However, for example, when processing a resin having a PET carrier film (PET carrier film), Wpf is required to be 70W. Here, if Wpf is required to be 70W, the output of the oscillator 1 in fig. 6 needs to be 500W, for example.

The laser oscillator described in the above example has an output characteristic such that a basic wave pulse waveform (output curve C in fig. 1, output curve a1 in fig. 11) having a1 st peak output smaller than a2 nd peak output at the time of stopping RF application is output immediately after RF application, but may be applied to a laser oscillator having an output characteristic such that a basic wave pulse waveform having a1 st peak output larger than a2 nd peak output at the time of stopping RF application is output immediately after RF application.

In summary, the laser processing method and the laser processing apparatus for printed circuit board of the present invention can achieve the object of the present invention.

The above description is only an example of the present invention, and the scope of the present invention should not be limited thereby, and the invention is still within the scope of the present invention by simple equivalent changes and modifications made according to the claims and the contents of the specification.

Claims (5)

1. A laser processing method of a printed circuit board, which performs laser processing using a carbon dioxide laser oscillator that oscillates laser light by applying an RF pulse, characterized in that:

during the laser output period after the application of the RF pulse is completed, the laser oscillation is continued by applying the RF pulse again, and the laser beam of a desired time is extracted from the continuously oscillated laser beam to perform laser processing of the printed circuit board.

2. The laser processing method of a printed circuit board according to claim 1, characterized in that: the application time of the RF pulse and the stop time of the RF pulse are controlled to generate a sawtooth pulse, and the sawtooth pulse is generated to perform machining.

3. The laser processing method of a printed circuit board according to claim 2, characterized in that: controlling at least one of a sum of an application time of the RF pulse and a stop time of the RF pulse, or a ratio of the application time of the RF pulse and the stop time of the RF pulse.

4. A laser processing apparatus for a printed circuit board, which performs laser processing using a carbon dioxide laser oscillator for oscillating a laser by applying an RF pulse, the laser processing apparatus comprising a control device, characterized in that:

the control device performs the following control: during the laser output period after the application of the RF pulse is completed, the laser oscillation is continued by applying the RF pulse again, and the laser beam of a desired time is extracted from the continuously oscillated laser beam and supplied to the processing unit.

5. The laser processing apparatus of a printed circuit board according to claim 4, characterized in that: the machining head laser control device may further include a plurality of machining heads configured to distribute laser light from the carbon dioxide laser oscillator to each of the machining heads, and the control device may supply laser light to one of the machining heads regardless of a positioning state of the other machining head when positioning of the one of the machining heads is completed.

Images