JPS63232344A - Method and apparatus for wire bonding - Google Patents

Abstract

PURPOSE:To obtain the stable loop height and form of a bonding wire by a method wherein the relative discrepancies between two bonding positions are detected and ideal bonding conditions for the respective lengths of connected wires are calculated in accordance with the detecting results to carry out wire bonding. CONSTITUTION:Relative discrepancies between 1st bonding positions 17-28 and 2nd bonding positions 17'-28' are detected and ideal bonding conditions for the respective lengths of connected wires are calculated in accordance with the detected results by a predetermined calculation method. With this constitution, a time required for re-teaching of the ideal bonding conditions which is necessary when the type of wire-bonding objects is changed can be substantially reduced and, moreover, the danger of misinput of various bonding conditions caused by an operator can be reduced and the danger of manufacturing defective products caused by the misinput can be also reduced as a result and, further, the degradation of the yield can be avoided.

Description

[Detailed description of the invention] [Purpose of the invention] (Industrial use) The present invention relates to wire bonding technology.

(Prior art) In the conventional wire bonding method of connecting multiple electrodes on a semiconductor pellet to multiple lead terminals using wires, ideal bonding conditions for each wire length are set in advance by an operator. It had been. However, when there are many types of wire-bonded objects, if the type of wire-bonded objects is changed during operation, the length of each wire to be connected must also be changed. Therefore, it was necessary for the operator to set the ideal bonding conditions for each wire length again.

Furthermore, in the conventional wire bonding technique, the position of the pellet to be mounted may deviate from the predetermined mounting position of the initially taught pellet (see FIGS. 4 and 5). In such a case, it is necessary to perform bonding using the initially taught ideal bonding conditions without modifying the ideal bonding conditions (conditions related to two strokes) for each wire length.

(Problems to be Solved by the Invention) Conventional bonding techniques have the following problems.

First, every time the type of the 1st K1 wire-bonded object was changed, a large amount of time was spent on teaching to reset the ideal bonding conditions for each wire length connected to each wire-bonded object. be.

Second, in relation to the first problem above, when resetting the ideal bonding conditions, there is a risk of erroneous input by the operator, and there is also a risk of manufacturing defective products due to this erroneous input. This also caused a decrease in yield.

Third, if the position of the pellet to be mounted is shifted from the predetermined mounting position that was initially taught, the initially taught bonding condition can be adjusted without modifying the ideal bonding conditions for each wire length. Bonding had to be carried out under the same conditions, and a stable loop height and shape could not be obtained.

[Structure of the invention]

(Means for Solving the Problems) In order to solve the above problems, the present invention detects a relative positional shift between the first bonding location and the second bonding location, and connects the wires based on the detection result. It is an object of the present invention to provide a wire bonding method and apparatus characterized in that wire bonding is performed by calculating ideal bonding conditions using a predetermined calculation method according to each wire length.

(Function) - According to the wire bonding method and apparatus as described above, the relative positional deviation between the first bonding location and the second bonding location is detected, and based on the detection result, the wire bonding method and device are configured to Since the ideal bonding conditions are calculated using a predetermined calculation method, it is possible to obtain a bonding body in which the connected wires have a stable loop height and shape, and it is also possible to improve the yield.

(Example) Hereinafter, one embodiment of the present invention will be described based on the drawings.

FIG. 1 is a schematic side view showing a wire bonding apparatus according to an embodiment of the present invention. In FIG. 1, (1) is a bonding head, which is placed on an X-Y table (2).
It is movable in the X-Y direction. The bonding head (1) has an internal rotating screw (not shown) for vertical movement for biaxial movement, and a vertical movement screw (not shown) for biaxial movement. A motor (3), which is a power source, is connected and mounted. A vertically movable body (4) is provided in front of the vertically movable screw, and a bonding arm (5) that is connected to an ultrasonic vibration source and performs ultrasonic vibration is further provided in front of the vertically movable body (4). A capillary (7) serving as a bonding tool is attached to the tip of the bonding arm (5), through which a wire (6) used for bonding is passed. Furthermore, a torch (8) is provided at the bottom of the bonding arm (5) for discharging and heating the tip of the wire (6) to form a ball. Further, a clamper (9) for cutting the wire (6) is provided on the upper part of the bonding arm (5).
is located.

On the other hand, a bonding stage (1G) is provided in front of the bonding head (1), and a vent αυ
(One board is mounted. Also, the above board αa
Above are the wire-bonded body, for example, the bullet round, and the bullet (11) to be bonded.
A positional deviation detection device (13) for detecting positional deviation is arranged so as to detect the positional deviation in a direction perpendicular to the pellet αυ and the substrate a, and is fixed to the support arm α.

In addition, the output from the positional deviation detection device αken is the A/D
Ideal bonder 4 before bonding via converter (102)
CPU (Central Pr
Abbreviation for cessinpUnit: central control unit) is sent to Kasumi.

Next, the operation of the wire bonding apparatus having the above structure will be explained with reference to FIGS. 1 to 3.

First, the substrate (the beret fixed to the I) (11) and the substrate (L
The relative positional deviation between the leads (IQ, . . .
From 1), a digitized signal is transmitted to the CPUQ5 via an A/D converter (102) as position correction information. After that, the signal (position correction information) sent to the CPU α91C
and the CPU (the storage unit (105) in the US).
The standard bonding history information (
From the new wire length for each positional deviation amount, and from this new wire length and the standard bonding conditions (conditions regarding Z stroke) that have been previously taught in the memory unit (105) in the CPU (Is). Ideal bonding conditions (
Conditions related to 2 strokes after correction) are compared to the comparison calculation section (1
03) Further, the correction calculation section (104) calculates and sets the result, and sends it to the calculation processing section (106).

After that, the CPU (I!19f) arithmetic processing unit (106)
Based on the ideal bonding conditions, the amount of movement of the pillar (7) in two axial directions is transmitted as an electrical signal to the bonding control unit (107) and further to the motor (3). On the other hand, the X-Y table (2) is cpU[! 9o Arithmetic processing unit (1o6) Color X-Y table control unit (108) K Operates based on the new information sent, so that the bonding head (1) can control the capillary (7) provided at its front tip. The first bonding point αη, αI, (19° (a), c2η, (2), +
23. CI! 4), (To), (A), Punishment, (To) is moved to an upper position, and then the tip of the capillary (7) is touched by a torch (8) by a signal from the bonding control section (107). A ball is formed, and then the rotation screw for vertical movement is rotated by the rotation of the motor (3) based on ideal bonding conditions, and the capillary (7) is lowered by the operation of the vertical movement screw, and a ball is formed at its tip. The balls are at the plurality of first bonding points αη.

α8. αs, (to), Q machine a, (23), h, (to), (c), □□□, (c) are pressed to a predetermined point from O°, and then the bonding arm (5) Bonding is performed by applying ultrasonic vibrations. The new information referred to here refers to information necessary to operate the present apparatus that has been corrected in accordance with the corrected ideal bonding conditions.

Next, the capillary (7) is raised by the operation of the vertical screw based on the ideal bonding conditions, and furthermore, based on the new information from the X-Y table control section (108)
Read αe by moving the y table (2).

...Second bonding point αη, bet, αs, m,
(2υ, @, (c).

do. After that, the capillary (
7) The motor (3
) rotates, and the rotating screw for vertical movement also rotates.
By operating the vertical screw, the force is lowered and the wire (6) is pressed, and bonding is performed by applying ultrasonic vibration of the bonding arm (5), and then the wire (6) is pressed by the clamper (9).
), the capillary (7) rises a little and then the clamper (9) rises, causing the wire (
6) is cut. Next, the capillary (7)
The Z-axis is moved from the X-Y table (2) K, which moves based on new information from the table control unit (108), to the third bonding point on the pellet αυ, and the capillary height during bonding is similarly corrected. Bonding is performed based on the amount of movement in the direction.

Next, one calculation method for determining the above-mentioned ideal bonding conditions will be explained using FIGS. 4 to 7. In Figures 4 and 5, the first bonding point (1η (111D
(1! (2G C2υ @ (ha) C41 (ha) (ha)
The pre-taught wire length between the lead 2!19 of the board α2 and the wire @ (to) is i5 = j,
(1=17.18...,28), and the capillary (7) in the biaxial direction corresponds to each J at this time.
Zx stroke Hzsti (1=17e IL”
'+28). Further, the values of 'l and Hzsti are stored in advance in the CPU (1!9).

In FIGS. 5 and 7, the length of the wire that is actually bonded after the position is recognized by the positional deviation detection device is 7? =j't (i=+17.18...,28), and the Z stroke of the capillary (7) in the Z-axis direction corresponding to each J/ at this time is H'zst i (
i = 17.18e...

28).

From FIG. 6, the trajectory of the capillary (7) taught in advance is drawn as A 4 B-+C, and the loop shape at that time is A + H-) C.

Here, A is Benz) The bonding point on (Ll),
B is the highest rising point of the lower end of the capillary (7), C is the lead α
The bonding points above Q, . . . and HJoop represent the loop height.

When lead length 1717 = 't〒, H2st17 + Hjoop + Jτ+(Hjoo
p+Hp)"-)-e-(7), and when lead length T7=11, Hzsti *Htoop + J-"(+ (Hjo
op + Hp)” + t fi)−7.

When the lead length is 2828 = J function, Hzst2B
Medium Hjoop + J",, + (HJoop +
Hp)"+t-6>), where Hp represents the thickness of the pellet αυ and e is a correction term.

From Fig. 7, the capillary (7) is actually bonded after its position is detected by the positional deviation detection device.
The locus is drawn as A'→B'→d, and the loop shape at this time is X→-→C'. Here, A', B',
C', H' represent the respective positions when the above persons B, C, - are actually bonded.

Lead length 17"17'=J-1,;1□,+ΔJ1.(
ΔJ-1, beam length correction amount), then g = ko J ′ I (HJoop + during zstry
+ t - (7), 1 lead length i #i "'
When =1'1=JI+Δzi, g vote=■J' Hzsti * Hjωp + +
g (1), lead length 28'28" =
When <” ” ””+ΔJ, C possible = -1・Iuoop + during H2st2S
+M (9). Here, equation (7) and (
From the formula 7), Hzst17*Hjoop +n possible Hjo
op + Hp)” + g, and similarly, formula (A) and 0
), (b) and Shio), the ideal bonding conditions for actual bonding can be found. Here, ΔJ is determined by the positional deviation detection device Sumi 1, and H
The values of IO and Hp are stored in the CPU before bonding in accordance with the object to be bonded and the bonding conditions.

According to the above method, bonding can be performed while maintaining a stable loop height and loop shape by controlling the two strokes of the capillary (7).

Here, when detecting the positional deviation by the positional deviation detection device Q3), each bonding point on the pellet αυ and each bonding point on the leads (11, . . . ) are provided regularly as shown in FIG. If it is possible to measure how much each of the bonding points ←η, (175, (c), and Δ) deviates from the predetermined position, then the degree of positional deviation of the remaining bonding points can be measured without measuring the degree of positional deviation of the remaining bonding points. Good.However, it is necessary to store in advance the position of each bonding point when no positional deviation occurs.

In addition, the positional deviation inspection device α= can measure the deviation between the preset bonding position and the actual bonding position, and can calculate the difference between the wire length at the time of setting and the actual wire length. It can be a means.

In addition, as a bonding method, you can measure the positional deviation of all bonding points at once and perform new bonding after setting the bonding conditions, or you can perform new bonding each time before bonding one wire. Various methods can be considered, such as a method of performing bonding by setting conditions.

〔Effect of the invention〕

According to the present invention, it is possible to significantly reduce the time required for re-teaching ideal bonding conditions when changing the type of wire-bonded object, and also to reduce the risk of incorrect input of various bonding conditions by the operator. As a result, the risk of manufacturing defective products due to incorrect input is reduced, and a decrease in yield can also be prevented.

Furthermore, even if the mounted pellet deviates from the predetermined mounting position that was initially taught, the ideal bonding conditions can be automatically corrected, making it possible to bond while maintaining a stable loop height and loop shape. .

[Brief explanation of drawings]

Fig. 1 is a side view of a wire bonding apparatus according to an embodiment of the present invention, Fig. 2 is a block diagram of the control system of the apparatus shown in Fig. 1, and Fig. 3 shows the operation of the apparatus shown in Fig. 1. A schematic flowchart, Figure 4 shows the bonding gear to be taught.
Figure 5 is a diagram showing the bonding locator 1 that is actually bonded, Figure 6 is a state diagram showing the bonding conditions that are taught, and Figure 7 is the bonding conditions that are actually bonded. FIG. (1)...Bonding head, (2)...X-Y
table. (3)...Motor, (4)...Vertical moving body. (5)...Bonding arm, (6)...Wire. (Force...cabin 2, (8)...torch. (9)...clamper, an...
bonding stage. αυ...pellet, αt...substrate. 0... positional deviation detection device, (14)... support arm. α9...CPU, (102)...
A/D converter. Agent Patent Attorney Noriyuki Ken Yudo Takehana Kikuo % review f/12 Figure 1! s 2 Figure Hayabusa 4 Figure 5 Figure 6

Claims (3)

[Claims] (1) In a wire bonding method in which a plurality of first bonding locations on a wire-bonded object and a plurality of second bonding locations corresponding to the plurality of first bonding locations are connected by wire, the first bonding location is wire bonding is performed by detecting the relative positional deviation amount between the second bonding point and the second bonding point, and calculating ideal bonding conditions using a predetermined calculation method for each wire length to be connected based on the detection result. A wire bonding method characterized by: (2) The wire bonding method according to claim 1, wherein the ideal bonding condition consists of displacement of the bonding tool in the vertical axis direction. (3) In a wire bonding apparatus that connects a plurality of first bonding points on a wire-bonded object and a plurality of second bonding points corresponding to the plurality of first bonding points with wires, the wire-bonding object positional deviation detection means for detecting the positions of the first and second bonding points, and ideal bonding conditions are calculated according to the length of each wire to be connected based on the signals from the positional deviation detection means, and bonding is performed. A wire bonding device characterized by having a control means for controlling operation.