JPH041863B2 - - Google Patents

Description

[Detailed Description of the Invention] [Field of Application of the Invention] The present invention relates to an inspection method for heating an object to be inspected having a bonded portion and determining the presence or absence of a bonding defect based on the temperature increase state of the heated portion, and the method for the same inspection. It is related to the device.

[Background of the invention]

An example of an object to be inspected to which this type of inspection method and apparatus is applied is shown in FIG. Figure a shows the soldering part of a flat package component, and figure b shows the wire bonding part of an LSI or the like.

Figure c is an explanatory diagram schematically showing the joint,
4 and 5 are members to be joined, 5 is a joining surface, 3 is member 4
This is the surface opposite to the joint surface in .

The bonding defects in such bonded products can be roughly classified into those where the bonded parts are completely separated, those where the bonded parts are only in contact, and those where the bonded parts are misaligned. In particular, among these defects, those where the joints are completely separated or those where the joints are only in contact are not only difficult to visually inspect, but also difficult to automate the inspection.

Known techniques for detecting such defects include testing by heating the soldered portion with a laser and measuring changes in the radiant temperature. Normally, the thermal resistance of the joint is low and the temperature rise is small, but if soldering is not done or is insufficient, the thermal resistance of the joint is large and the temperature of the heated part rises rapidly. . Vanzetti's method applying this principle (US Patent No. 3803413)
There is. The disadvantages of this method are as follows. Laser energy absorption and thermal emissivity vary greatly due to differences in surface conditions and surface inclinations. It is impossible to distinguish between good and defective products using only the maximum value and rate of change of the detected radiation temperature. Can not. Therefore, in the above-mentioned Vanzetsutei method, a large amount of samples are subjected to an inspection device in advance to create statistical data on the maximum temperature and temperature increase rate for each inspection site. Based on the data, defect determination criteria are created for each inspection site. Inspection is performed by comparing the maximum temperature and temperature rise rate of the object to be inspected with defect criteria for each part. but,
Even if this method is used, it is not possible to sufficiently remove the effects of surface conditions and surface inclinations, and in addition,
A great deal of time and effort must be spent preparing for the test.

[Purpose of the invention]

The present invention has been made in view of the above circumstances, and eliminates the need to spend a great deal of time and effort in preparing for inspection.Moreover, it automatically corrects the influence of the surface condition and surface inclination of the object to be inspected. Moreover, it is an object of the present invention to provide an inspection method that can reliably determine the presence or absence of a bonding defect, and an inspection apparatus suitable for implementing the above method.

[Summary of the invention]

In order to achieve the above object, the present invention provides an inspection method, that is, in order to achieve the above object, the present invention heats an object to be inspected having a joint, and uses heat radiation radiated from the heated portion. detect and measure the temporal change in radiant temperature, and detect the surface condition of the object to be inspected based on the radiant temperature change measured during a short period of time that is not affected by defects in the bonding area. A correction based on at least the detected surface condition of the object to be inspected is added to the radiation temperature change measured over a relatively long period of time affected by the defect, and based on this corrected radiation temperature change, This is a bonding state inspection method characterized by determining the presence or absence of defects at a bonding site. The present invention also provides a heating means for heating an object to be inspected having a joint, and a detecting means for detecting thermal radiation of a heated part heated by the heating means, and detecting the radiation temperature based on the thermal radiation over time. a measurement means for measuring the change;
surface state determining means for calculating a correction coefficient indicating the surface state of the object to be inspected based on a radiation temperature change measured by at least the measuring means during a short period of time that is not affected by defects in the joint portion; , a correction coefficient calculated by the surface condition determining means to correct the radiation temperature change measured by the measuring means over a relatively long period of time, which is affected by the presence or absence of a defect in the joint portion; A bonded state characterized by comprising a bracing means and a defect determining means for determining the presence or absence of a defect in a joint based on a radiation temperature change over a relatively long period corrected by the calibration means. This is an inspection device.

Next, the principle of the present invention will be explained with reference to FIG. This figure is an explanatory diagram showing the relationship between heat flow and time when one of two joined members is heated.

4 and 6 are faceted members, and 5 represents the thermal resistance at the joint surface. This thermal resistance varies depending on the bonding state, and if the bonding is incomplete, the thermal resistance increases.

For this reason, if heat is applied to the surface 3 of the member 4 as shown by arrow A and the surface temperature is measured by radiation from the surface 3 (arrow B), the presence or absence of a bonding defect can be estimated from the temperature increase state. That is, if the bonding is incomplete, the thermal resistance 5 is large, and therefore the member 4 accumulates heat and rapidly rises in temperature. In the prior art, the above principle was used to determine whether the bonding state was good or not based on the temperature rise state of the surface 3.

Generally, it takes a certain amount of time for heat to be conducted through an object. Therefore, as shown in FIG. 2, in order to know the state of a place δ away from the heated part from the heated part, a time t _a that is twice the time required for heat to be conducted by δ is required. When inspecting a joint using heat, assume that the heated part (surface 3 in the example in Figure 2) is the same as the radiation temperature measurement location, and that there is a distance δ from the heated part to the joint 5. Ba,
At time t where O<t<t _a , a radiation temperature determined by the surface state of the heated region is observed, regardless of the state of the joint. Further, at time t where t _a <t, a radiation temperature determined by the surface state and the joint state is observed.

Using the above phenomenon, if we correct this based on the temperature change observed after time t _a and further using the temperature change observed before time t _a , we get:
It is possible to accurately estimate the bonding state by removing the effects of surface conditions and surface inclinations.

[Embodiments of the invention]

Next, one embodiment of the present invention will be described with reference to FIGS. 3 to 8.

FIG. 3 shows the object to be inspected in this example,
A pad 8 is provided on the substrate 7, and a lead 9 is soldered to the pad 8.

In this embodiment, the explanation will be limited to the inspection of soldered parts, but the inspection can be similarly carried out on objects having the structure shown in FIG. 1C.

FIG. 4 is a schematic perspective view showing an inspection apparatus of the present invention configured to carry out the inspection method of the present invention.

A heating section 12 is provided to heat the inspection section 11 of the soldered lead 9 for a certain period of time. In this embodiment, the main heating section 12 is constructed of a YAG laser, but when implementing the present invention, any heating means such as a carbon dioxide laser can be used.

An illumination system consisting of an irradiation optical system 14 for irradiating the inspection object with laser light, and a wavelength λ of InAs, HgGaTe, InSb, PbS, etc. of 0.8 μm to 15 μm for measuring the radiation temperature of the inspection object surface.
a detection optical system 16 using a Cassegrain system or an off-axis system or a lens for guiding the thermal radiation emitted from the surface of the object to the detector; A detection system including a filter 17 for preventing light from entering the detector, an X-Y table 18 and a control unit 19 for sequentially scanning the detection system and the object to be inspected are provided. A block diagram of the control section 19 is shown in FIG. The control unit 19 includes an X-Y table control unit 20 that drives an X-Y table, and a heating control unit 2 that controls heating of the scanned component by opening and closing the shutter 13.
1. An amplifier 22 that amplifies the measured temperature signal converted into an electrical signal from the detection section; A/D converter 2 that converts the amplified analog signal into a digital signal;
3. Sample clock 24, which generates the clock that determines the timing of A/D conversion;
A surface condition determination unit 25 receives the temperature within a certain period of time from the start of heating using a trigger from the clock, determines the slope and surface condition of the surface, and calculates a correction coefficient C, which will be described later. It is composed of a calibration section 26 that corrects the temperature, a defect determination section 27 that determines and displays defects using the corrected maximum temperature of the measured temperature and temperature change rate, and an overall control section 28 that controls the entire system. There is. The above-mentioned surface state determination unit 2 in this embodiment
5. Calibration section 26, defect determination section 27
represents dedicated hardware or software such as a microcomputer, and the overall control section 28 is constituted by a computer such as a microcomputer or a minicomputer.

When using the inspection apparatus configured as described above, standard samples of non-defective products and defective products are selected in advance, and the results of inspection operations performed on each sample are stored in the memory of the overall control section 28.

Next, set a short time t _a that is not affected by the condition of the soldered part. The following method is recommended as an example of how to set this time t _a . That is, for each of the above-mentioned non-defective sample and defective sample, several types of temperature increase conditions were measured for an extremely short period of time to a relatively long period of time, and a value as large as possible was taken as long as the data on both did not show a significant difference. Let it be t _a .

Further, the time t _b for measuring the temperature increase state is determined as follows.

Let t _c be the minimum time during which structural members in the vicinity of the object to be inspected are affected by heat or are likely to be affected by heat by heating the object to be inspected.

The test duration t _b is set within the range of t _a << t _b << t _c .

Before starting inspection operations, first prepare as follows. The overall control unit 28 moves the X-Y table 18 to the initial position as initialization,
The shutter 13 is closed, the heating source 12 is turned on, and the sample clock 24 is reset.

Next, the X-Y table 18 is driven by a command from the X-Y table control section 20 to move the object to be measured to the inspection position. After the object to be measured is moved to the inspection position, the sample clock 24 is reset and measurement of the radiation temperature of the heating section 11 to be inspected is started. A little later, the shutter 13 is opened and heating begins.

The measured radiation temperature is passed through the A/D converter 23 in accordance with the timing sent from the sample clock 24 and digitized. Of the digitalized radiation temperature, the measured temperature data is taken into the surface state determining section 25 until the time t _a when the temperature changes without being affected by the state of the soldered part. Surface condition determination unit 25
calculates a correction coefficient C based on the captured data using a method described later, and sends the correction coefficient C to the calibration section 26. The calibration section 26 multiplies the measured temperature data by a correction coefficient C to create corrected temperature data and sends it to the defect determination section 27 . If the corrected temperature data sent is significantly different from a normal lead, the defect determining section 27 determines this as a defect and displays the defect.

After the time t _b required for determination has elapsed, the overall control unit 2
8, the sample clock 24 is reset and the X-Y table 18 is driven to move the next measurement object to the inspection position. Thereafter, the same operation is repeated until all the measurement targets are inspected and the inspection of one board is completed.

Next, the surface condition determination section 25, calibration section 26, and defect determination section 27 will be explained in more detail.

First, the surface condition determination unit 25 determines the time period from the start of heating, which causes a temperature change that does not depend on the condition of the soldered portion.
At least two digital radiant temperatures are captured before t _a has elapsed. The reason for capturing two or more points is as follows. (a) It is necessary to measure the temperature immediately before heating in order to calibrate the temperature of the object to be inspected on the heating surface and the state of the atmosphere in which the object is contained. (b) At least one point is required to calibrate the difference in absorption rate and thermal emissivity due to the inclination and surface condition of the surface of the heating section.
(c) When the object is extremely small or the optical system
When a sufficient NA cannot be obtained, the S/N ratio deteriorates as shown by the following equation, and the variation in error becomes extremely large if only one point is used.

S/N=α・(NA) ²・√sample interval…(1) where α: constant of proportionality Therefore, if the above-mentioned radiant temperature is captured at two or more points, calibration can be performed accurately using a method such as the least squares method. can be done.

In addition, the temperature changes in a mode determined only by the inclination and surface condition of the surface of the heating part, and the time _{t a} for temperature change without being affected by the condition of the soldering part.
For example, a lead with a thickness of 0.2 mm and a width of 0.5 mm is
When soldering is performed on a pattern with a width of 0.7 mm and a width of 0.7 mm, t _a ≒0.3 ms. The evaluation of this value is
It can also be calculated using the following formula for one-dimensional step heating.

t _a =2√12・Thickness (K: temperature conductivity) ...(2) As shown in Figure 6, the radiation temperatures 29 taken in as described above are sequentially designated as T ₀ , T ₁ , ...Tn .
(Here, T ₀ is the temperature just before heating, T ₁ , . . . Tn are the temperatures when each sample clock occurs in order, or the average temperature between samples, and it is not necessary to sample at equal intervals. ) Measure the radiant temperature of a predetermined standard sample of good quality products at similar time intervals.

Figure 7 shows the radiation temperatures T ₀ °, T ₁ °, and
T ₂゜...Tn゜ and the radiation temperature of the object to be inspected T ₀ , T ₁ , T ₂
...This is a chart comparing Tn (see Figure 6).

In this way, the temperature curve 30 of the good sample
The difference between the temperature curve 29 and the temperature curve 29 of the object to be inspected is due to differences in the surface condition and surface inclination at the initial stage of measurement, and during the minute time span up to the aforementioned time t _a It doesn't show whether it's good or bad.

By correcting the temperature difference due to the difference in surface condition and surface inclination described above based on the temperature curves 29 and 30 between time t ₀ and _{t o} , a temperature curve 31 is corrected as shown in FIG. When this corrected temperature curve 31 and the temperature curve 30 of the non-defective sample are superimposed, it becomes as shown in FIG. 8, and they almost match between times t ₀ and t ₀ . Furthermore, in order to perform the above superposition operation computationally, a correction coefficient C for performing conversion between both curves 29 and 30 is determined. To find the correction coefficient C, there are the least squares method, a method of comparing sums, and a method of comparing weighted sums, and the correction coefficient C is found by any of these methods.

Using the least squares method, the correction coefficient is expressed by the following equation.

C=Σ(Sk ₂ )/ΣSk(Tk−T ₀ )...(3) Next, the calibration section 26 calculates the coefficient representing the slope and surface condition of the surface of the heating section 11 determined by the surface condition determining section 25. The correction coefficient C and the digitized radiation temperature from the A/D converter 23 are
Corrected radiant temperature Tn calculated from Tn ₊₁ , Tn ₊₂ , ..., Tm (t _a < t ≤ t _b , temperature when each clock occurs or integral value from the previous clock) using the following formula: Calculate ′ ₊₁ , Tn′ ₊₂ ,…, Tm′.

T _k ′=C(T _k −T ₀ )(k=n+1, n+2,...,
m) …(4) This corrected radiation temperature Tn′ ₊₁ , Tn′ ₊₂ ,…,
Tm' and the correction coefficient C are sent to the defect determination section 27.

In the defect determination section 27, the calibration section 26
Correction coefficient C sent by and corrected radiant temperature
Defects are determined from Tn′ ₊₁ , Tn′ ₊₂ , ..., Tm′.

The corrected maximum temperature Tm'ax of the radiation temperature and the temperature difference ΔT _kl = T _l '−T _k ' between the sample timings commanded by the overall control unit 28 are determined.

(n+1≦k<l≦m) Here, the sample timing for calculating the temperature difference is to input good products and defective products several times in advance, and find a timing that allows easy separation of good products and defective products. . When the radiation temperature from the object to be inspected is detected and input to the control unit 19,
By performing the above calculation, if the following formula Ca<C<Cb… Ta<Tm′ax<Tb… ΔTa<ΔTkl<ΔTb… (5) is satisfied, it is determined to be a good product, and the condition of this formula is If the requirements are not met, the product will be judged as defective. In addition, Ca
The reason for determining <C<Cb is that the surface condition of the object to be inspected is significantly different from that of a non-defective item and is abnormal, so that it is determined to be a defective item. Here, Ca, Cb, Ta, Tb, ΔTa, and ΔTb are inputted several times in advance for non-defective products and defective products, and the rate of determining defective products as non-defective products is extremely low. Set to a low value.

These operations speed up the inspection of soldered joints on flat package components, especially for floating lead defects (including those that are completely floating and those that are in contact but are not soldered). Tests can be performed with high reliability.

The above has described an example using the least squares method based on equation (3), but as an example different from the above, the correction coefficient C is calculated by comparing the sums as shown in equation (6) below. You can also ask for it.

C=ΣS _k /Σ(Tk−T ₀ ) …(6) In this method, if the sum of standard temperature changes is memorized, the sum or integral of the captured temperature changes and the
It is possible to obtain the correction coefficient by only dividing the number of times, and it is a simple and fast method. In the above example, the case where the correction coefficient C was calculated using the temperature curve of the non-defective sample and the temperature curve of the object to be inspected was explained. The correction coefficient C' is obtained using the temperature curve of the object to be inspected, and the corrected radiant temperature Tk' can be calculated using equation (4) using this correction coefficient C'.

That is, in equation (6), ΣSk is the sum of the relative temperatures of non-defective samples, can be determined through experiments, etc., and can be regarded as a constant in practice.

Therefore, by dividing the correction coefficient C in equation (6) by this constant and transforming it, the correction coefficient C' can be obtained from equation (7).

C′=C/ΣSk=1/Σ(Tk−T ₀ )…(7) The correction coefficient C′ transformed in this way and the corrected temperature Tk′ calculated by it are simply divided by a constant. Therefore, the defect can be determined in the same manner as in equation (5).

Further, in the comparison of weighted sums, the correction coefficient C is obtained by the following equation.

C=Σa _k S _K /Σa _k (T _k −T ₀ ) (8) where a _k is a predetermined weight.

Equation (8) above is a method that is an extension of the least squares method, and according to this equation, it is possible to obtain a more accurate correction coefficient than the other methods described above.

FIG. 9 shows an embodiment different from the above-mentioned FIG. 5,
The A/D converter 23, the surface state determination section 25, the calibration section 26, and the defect determination section 27 are respectively replaced by a relative temperature calculation section 32, a storage section 33, an integration section 34, an integration section 35, a constant division section 36, and a constant. Multiplication section 3
7. By replacing the correction coefficient comparison section 38, maximum temperature calculation/comparison section 39, and temperature difference calculation/comparison section 40, it is possible to handle everything with analog quantities.

For example, when using the calibration method and the least squares method, the following calculations are performed for each part.

The relative temperature calculation unit 32 calculates the relative temperature Trel(t)=T _(t) −T ₍₀₎ based on the temperature immediately before heating, and the storage unit 33 stores the standard temperature change T゜(t) and ∫ of a non-defective product. ^ta ₀
Remember T゜(t) ² dt (instead of Σ(T゜_k ) ² ),
The constant integration unit 34 calculates the product of the relative temperature Trel(t) and the stored standard temperature change T゜(t), and the integration unit 35 calculates _the product of the relative temperature _Trel (t) and the stored standard ^{temperature change} T゜(t). (T _k −T ₀ )・T゜
(instead of _k ), and the division unit 36 calculates C=∫ ^ta ₀ (T゜(t)
² dt/∫ ^ta ₀ (T(t)-T(0)・T゜(t)dt is calculated to obtain the correction coefficient, and the constant multiplier 37 calculates T'(t)=
Calculate C·Trel(t), and a correction coefficient comparison unit 38;
Maximum temperature calculation/comparison section 39, temperature difference calculation/comparison section 40
is the part that calculates each quantity and determines defects using equation (5). In this embodiment, since all signals are handled as analog quantities, the main part can be constructed from an operational amplifier and one analog storage section, resulting in an inexpensive and high-speed control section.

Furthermore, if the calibration method is performed by comparing the sums and formula (7) is used, the surface state determining section 25 can be further simplified as shown in FIG. can do. The integration section 41 calculates ∫ ^ta ₀ T(t)dt, and the reciprocal calculation section 42 calculates 1/∫ ^ta ₀ T(t)dt.

In this embodiment, the main part is composed of only operational amplifiers, resulting in a very inexpensive and simple device.

[Effects of the Invention] As detailed above, according to the inspection method of the present invention, there is no need to spend a lot of time and effort on inspection preparation, and the influence of the surface condition and surface inclination of the object to be inspected can be avoided. The presence or absence of a bonding defect can be determined quickly and reliably by automatically correcting it.

Moreover, according to the inspection apparatus of the present invention, the above-described inspection method can be easily implemented and its effects can be fully exhibited.

[Brief explanation of drawings]

FIGS. 1A, 1B, and 1C are perspective views showing an example of a joining member that is an object to be inspected in the present invention, and FIG. 2 is a schematic diagram for explaining the principle of the method of the present invention. 3 to 8 show an embodiment of the method of the present invention carried out using an example of the device of the present invention, FIG. 3 is a perspective view of an object to be inspected, and FIG. 4 is a schematic diagram of the inspection device. Figure 5 is a block diagram of the calculation part, Figure 6 is a perspective view,
FIG. 7 and FIG. 8 are charts showing temporal changes in radiation temperature, respectively. FIGS. 9 and 10 are block diagrams of inspection apparatuses according to embodiments different from those described above. 1... Soldering part of flat pack parts, 2
...Wire bonding part of LSI etc., 3...
Heating section and temperature measuring section, 4...objects 1, 5...
Junction of object 1 and object 2, 6...object 2, 7...
Board surface, 8... Wiring pattern on board surface, 9...
Lead, 10...Flat pack type electronic component,
11... Inspection section, 12... Heat source, 13... Shutter, 14... Heating section, 15... Detector, 1
6...detection optical system, 17...filter, 18...
X-Y table, 19...control section, 20
...X-Y table control section, 21 ... heating control section, 22 ... amplifier, 23 ... A/D converter, 2
4...Sample clock, 25...Surface condition determining section, 26...Calibration section, 27...
Defect determination section, 28... Overall control section, 29... Measured radiation temperature change, 30... Standard temperature change of non-defective product, 31... Corrected radiation temperature change, 32... Relative temperature calculation section, 33... Memory Part, 34... Integration part, 35... Integration part, 36... Constant division part, 37
...Constant multiplication section, 38...Correction coefficient comparison section, 39
...Maximum temperature calculation/comparison section, 40...Temperature difference calculation/comparison section, 41...Integration section, 42...Reciprocal calculation section.

Claims (1)

[Claims] 1. An object to be inspected that has a joint is heated, thermal radiation emitted from the heated part is detected, and the temporal change in the radiation temperature is measured to determine whether the influence is affected by a defect in the joint. The surface condition of the object to be inspected is detected based on the change in radiant temperature measured over a short period of time, which is not affected by the temperature change, and the surface condition of the object is detected based on the change in radiant temperature measured over a relatively long period of time, which is affected by defects in the joint area. A method for inspecting a bonded state, characterized in that a correction is made based on at least the detected surface condition of the object to be inspected, and the presence or absence of a defect in the bonded portion is determined based on the corrected radiation temperature change. 2. The bonding state inspection method according to claim 1, wherein the correction is further performed using a short-time radiant temperature change obtained from a non-defective inspection object. 3. A heating means for heating an object to be inspected having a joint, and a measurement in which a detecting means detects thermal radiation of a heated part heated by the heating means and measures temporal changes in radiation temperature based on the thermal radiation. and a surface condition for calculating a correction coefficient indicating the surface condition of the object to be inspected based on the radiation temperature change measured by at least the measuring means during a short period of time unaffected by defects in the joint part. a determining means, and a correction coefficient calculated by the surface condition determining means to correct a radiation temperature change measured by the measuring means over a relatively long period of time that is affected by a defect in the joint portion. A joint characterized by comprising: a calibration means; and a defect determination means for determining the presence or absence of a defect in a joint based on a radiation temperature change over a relatively long period of time corrected by the calibration means. Condition inspection equipment. 4. The surface condition determining means includes a storage means for storing the short-time radiant temperature change obtained from a non-defective inspection object, and the above-mentioned correction coefficient is further stored as the short-time radiant temperature change. The bonding state inspection device according to claim 3, characterized in that it is configured to calculate based on .