JPH0758314B2 - Circuit board inspection method - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a circuit board inspection method for inspecting the quality of a circuit board on which electronic components and the like are mounted, and more particularly to a positive half-wave of an AC signal for measurement. And a circuit board mounted with elements having different impedances for negative half-waves.

[Conventional example]

Circuit board inspection devices called in-circuit testers have come to be used for inspection of circuit boards on which electronic components and the like are mounted.

An example thereof is shown in FIG. 7. For example, a circuit board to be inspected (hereinafter referred to as “test board”) 3 which emits a measuring AC voltage having a predetermined frequency from a signal source 1 and which passes through an amplifier 2.
In addition, a current that is inversely proportional to the magnitude of the impedance Z _X flows through the substrate 3. This current is taken into and detected by, for example, the current detector 4, converted into a voltage here, then digitally converted by the A / D converter 5 and added to the measuring unit 6. The measurement unit 6 uses this variable data to calculate the impedance Z _X of the test board 3 and compares it with the reference value Z _S stored in advance in the memory. In this case, as shown in the following formula, if it is within the predetermined tolerance ± α, it is judged as good, and if it is outside the tolerance, it is judged as bad, and the judgment result is displayed on the display 7 or the like. There is.

Determining Z _S -αZ _X Z _S + α determined Z _X <Z _S -α or Z _X> Z _S + α [Problems to be Solved by the Invention] quality determination method in the conventional substrate inspection "bad" in the "good" Has the advantage of being simple. By the way, as is well known in a circuit board inspecting apparatus having a digital measuring system, for example, 1 LSB is added to the analog input voltage at the time of AD conversion.
There is an insensitive range that digital data does not change even if there is a considerable voltage change, and so-called quantization error occurs.

In this case, if the current flowing through the test board is relatively large, the input voltage of the A / D converter is also large, so this does not pose a problem.However, if the flowing current becomes small, the dead voltage increase in the input voltage of the A / D converter Becomes large and the measurement error increases, which cannot be ignored. However, since this point is not always taken into consideration in the conventional board inspection, for example, even if the value of Z _X is within the range of Z _S ± α, the measurement error due to the quantization error is added as the measurement value. There is a case where the range is deviated and "good" is erroneously determined as "bad".

The present invention has been made in view of the above circumstances, and an object thereof is to allow the measurement error component based on the quantization error to be a tolerance for the original variation of the measured impedance Z _X according to the size of the digitally converted data. It is an object of the present invention to provide a highly accurate circuit board inspection method that automatically adds up and prevents erroneous determination.

[Means for Solving the Problems]

Referring to FIG. 1, which illustrates an embodiment of the present invention,
In order to solve the above problems, the following means (a) to (c) are provided.

I. For example, for n circuit boards that have been confirmed to be non-defective in advance, an AC voltage for measurement of one cycle is applied from the signal source at each measurement point, and the average value of the current that flows during the + half-wave period and the − half-wave period. Impedance Z corresponding to the current average value in the + half-wave period and the − half-wave period
A reference data memory 20c and a measurement data memory 20d that determine μ _SA and Z μ _SB and hold these values as reference impedances Z _SA and Z _{SB at the} measurement point.

B. The root mean square σ is obtained from the current data and the average value thereof in the + half-wave period and the − half-wave period of each measurement point, and, for example, in the + half-wave period and the − half-wave period based on the root mean square value σ. Measured impedance Z _XA and Z _XB
And the tolerances α _1A and α _1B for the variation of the impedance are set, and the measurement error due to the quantization of the impedance Z _X , that is, the reading error is, for example, the tolerance related to the size of the digital conversion data of the measured value of Z _X. Difference α _2A ,
Tolerance setting means 20a for setting α _2B .

C. For example, in the impedance measurement in the + half-wave period and the − half-wave period, the tolerance α ₁ of the above variation and the reading error α ₂
Is automatically added, and impedance comparison is performed according to the following equation to determine whether the test board 3 is defective or not.

(Judgment formula) + Half-wave period Z _SA − (α _1A + α _2A ) Z _X Z _SA + (α _1A + α _2A ) ……
(1a) - half-wave period _{Z SB - (α 1B + α} 2B) Z X Z SB + (α 1B + α 2B) ......
(1b) [Operation] By including the above means, the quality of the test board is correctly determined even if the impedance measurement value includes a measurement error due to quantization.

〔Example〕

In this embodiment, first, a plurality of substrates that have been confirmed to be non-defective are measured in advance and the data are collected, and then a test substrate is measured by the same method, and then the data of both are compared to determine pass / fail. I am supposed to do it.

Referring to FIG. 2 together with FIG. 1 above, for example, a signal source
A one-cycle AC voltage for measurement as shown in FIG. 2A is emitted from 11 and applied to the non-defective substrate 3 via the amplifier 12. As a result, a current that is inversely proportional to the magnitude of the impedance Z _S flows through the substrate 3. in this case,
When a diode or the like is mounted on the substrate 3, the impedance is low for the positive half-wave of the measuring AC voltage and high for the negative half-wave, so the current flowing is shown in FIG. become that way. This current is detected by, for example, the current / voltage converter 13, and then converted into a voltage with a predetermined gain as shown in FIG.

In this embodiment, the operation of the signal source 11 and the current / voltage converter 13 is controlled by, for example, the measuring unit 20, and
Sends an AC voltage of a predetermined frequency for one cycle as described above. Further, the current / voltage converter 13 detects a current in the positive direction and a current in the negative direction which flow through the substrate, and converts them into positive or negative voltages of the same polarity with a constant gain. The switching controller 15, for example, shapes the positive half wave and the negative half wave of the measuring AC voltage emitted from the signal source 11 into square wave voltages, and outputs the waveform to switch the switch 14 to the second wave.
As shown in FIG. 4D, the positive half-wave period is driven to the contact A side, and the negative half-wave period is driven to the contact B side.

As a result, the converted voltage of the current flowing through the substrate 3 in one half-wave period 0 to π is integrated by the integrator 16 as shown in FIG.
The converted voltage at 2π is integrated by the integrator 17 as shown in FIG. The integrated voltage of these good boards is V
_{Assuming SA} and V _SB , these two integrated voltages are A / D converters 18 and 1 respectively, as shown in (g) and (h) of FIG.
Digitally converted at 9 and sent to the measuring unit 20. The measuring unit 20 uses the integrated voltage data to measure the center impedance of the non-defective board for each positive and negative half wave of the measuring AC voltage.
In addition to calculating Z _SA μ and Z _SB μ, the tolerances α _1A and α _1B for variations in the impedance of the test board are obtained by, for example, the tolerance setting means 20a, and at the same time, the tolerance addition α due to the quantization error α _After taking _2A and α _2B , impedances Z _XA and Z _XB in the positive half-wave and the negative half-wave of the test board 3 are measured by the same method as above, and the data are compared by the equations (1a) and (1b). After the
The judgment result of good or bad and the measurement data are sent to the recording / display unit 21 to be displayed and recorded.

In this case, the addition value α ₂ of the measurement error due to the quantization and the tolerance α ₁ with respect to the impedance variation in the expressions (1a) and (1b) is calculated by the following expression α ₁ = k ₁ σ + k ₂ Z μ [Ω] ... (2) α ₂ = Zμk ₃ ρ [Ω] ・ ・ ・ (3) In the above equation, k ₁ , k ₂ , and k ₃ are constants, and ρ is a variable related to the size of the digital conversion data of the positive and negative half waves.

Hereinafter, a method for obtaining the above α ₁ and α ₂ will be described. Now
Full-scale input voltage V _{FS with} N-bit A / D converter
When 2 is digitally converted, 2 ^N pieces of digital data are obtained. In this case, there is a difference of 1 digit in the smallest bit, that is, LSB, between adjacent data. It is well known that the difference of this magnitude is converted into the analog input voltage ΔV, which is represented by ΔV = V _FS × 2 ^−N (4), and the full-scale input voltage V _FS and the conversion bit number N
It is a constant value determined by.

Therefore, even if the converted analog voltage input to the A / D converter changes within the range of ΔV or less, the digital data does not change, and this voltage ΔV becomes the dead range of the A / D converter. The above 2- ^N is generally called the resolution of the A / D converter, and it is ± ΔV / 2 for an ideal A / D converter adjusted so that the digital data matches the median of the dead band voltage ΔV. The corresponding ± 1 LSB / 2 is called the quantization error.

Here, as shown in FIG.
The voltage lower than V _FS is V _SA μ, and its digital conversion data is
If D _SA , other voltage V within ΔV including this voltage V _SA μ
_SA , V _SA ′, etc. are also converted into data D _SA .

Therefore, if the true values of the impedances corresponding to the voltages V _SA μ, V _SA , V _SA ′, ΔV are D _SA μ, D _SA , D _SA ′, ΔD _A , then | D _SA -D _SA ′ | ΔD _A In the range of, naturally, D _SA = D _SA ′ = D _SA μ is measured, and a quantization error occurs.

Therefore, ε _A = | D _SA -D _SA μ | / D _SA μ or ε _A = | D _SA ′ −, where ε _A is the reading error based on digital data conversion, that is, the ratio of the quantization error to the data read value. D _SA μ | / D _SA μ, and when expressed in analog voltage, ε _A = | V _SA -V _SA μ | / V _SA μ or ε _A = V _SA ′ −V _SA μ / V _SA μ .

The maximum error is when the numerator on the right side of the above equation becomes equal to ΔV, so generally ε _A = ΔV / V _SA μ (5). In this case, since the numerator ΔV on the right side is a constant value as described in the explanation of the equation (4), the denominator V _SA μ may be increased in order to reduce the reading error.

By substituting equation (4) into equation (5) and rewriting it, ε _A = V _FS × 2 ^−N / V _SA μ Here, V _FS is the full-case input voltage, so the N-bit A / D converter The converted data corresponds to 2 ^N. Further, since V _SA μ similarly corresponds to D _SA μ, the above equation can be replaced with ε _A = 2 ^N × 2 ^−N / D _SA μ (5 ′).

Therefore, D _SA μ = 100 at the epsilon _A = 0.01 in the above equation (01100100) is obtained when the reading error epsilon _A example within 1%, A / D converter input voltage (K) is to satisfy this It should be set to. In this embodiment, for example, the measuring unit 20 monitors the conversion data (FIG. 2 (g)) of the A / D converter 18 and inputs it, that is, an integrator.
The current of 18 output (Fig. 2 (e)) becomes K /
The gain of the voltage converter 13 is controlled and fixed.

In FIG. 4 (a), assuming that the same measurement points of n substrates which have been confirmed as non-defective products are measured in analog, the average value V of the integrated voltage V _SA i during the half wave period 0 to π is V. _SA μ for example And an example of a case where the distribution of each integrated voltage with respect to this average value is plotted is shown. However, the horizontal axis of the figure represents the analog voltage, and the vertical axis represents the number of measurement data (the number of substrates). Here, when n is set to an appropriate size, the distribution state of the number of data is the average value as shown by the solid line, for example.
Approximate a normal distribution symmetrical to V _SA μ.

Therefore, the root mean square value σ _SA From the above, 99% or more of the number of test substrates n falls within the range of ± 3σ _SA , and the number outside this range is 1% or less as well known.

In this case, since all the tested substrates are non-defective, it is necessary to judge the number of 1% or less deviating from ± 3σ _SA as non-defective.

Here, when each analog voltage data V _SA i is digitally converted by an A / D converter having an extremely large number of bits and impedance is obtained so that a measurement error due to quantization can be ignored, for example, in (a) of FIG. As shown by the solid line, an impedance distribution diagram similar to FIG. 4 (a) is obtained. However, in practice, an 8-bit A / D converter is used, and a measurement error associated with 1% quantization, that is, a reading error is allowed. Therefore, impedance Z _SA i
The distribution of the reading D _SA i of the digital conversion data corresponding to is as shown by the dotted line in the figure.

So even considering measurement error etc. In this embodiment, for example, k ₁ to 3, k ₂ 2% of formula (2) Distant, when negligible reading errors, Therefore, α ₁ = 3σ _SA + 0.02Z _SA μ [Ω] …… (2a) or α ₁ = (3σ _SA / Z _SA μ + 0.02) × 100 [%] …… (2b) .

Next, the measurement error due to quantization, that is, the addition α ₂ to the tolerance for the reading error will be described. As shown in FIG. 2 above, the impedance is significantly different when a positive voltage is applied to the diode and when a negative voltage is applied, for example. Now, if the gain of the current / voltage converter 13 of FIG. 1 is controlled so that the + half-wave side digital conversion data ((g) in the figure) becomes approximately 100 counts (01100100), the − half-wave side digital The converted data (Fig. 2 (h)) has a value much less than 100 counts, and therefore its reading error becomes large.

When the variation in impedance on the half-wave side can be regarded as a normal distribution, the distribution can be plotted as shown in FIG. 4 (b), for example. However, when the digitally converted data of the voltage having such a distribution characteristic is plotted, when the converted data value is large, the solid line in FIG. As a result, the variation becomes large as indicated by the dotted line in the figure.

This increase in variability is originally but are intended to be incorporated into the alpha ₁ when the calculation above alpha _1, not completely necessarily all absorbed by such as when the number of samples n of the non-defective substrate is small,
It cannot be ignored.

In this case, the leading error ε _B has a center value of D _SB μ
Then, ε _B = 1 / D _SB μ …… (6) According to the equation (5 ′), this% is displayed and expressed using the variable ρ ε _B = ρ × 100 [%] …… (7) , Ρ = 1 / ( _DSB μ × 100).

Let this error be the tolerance α ₂ associated with the reading error,
When generalized, α ₂ = ε _B = ρ × 100 [%] (3a). For the central impedance Zμ, α ₂ = Zμ · k ₃ · ρ [Ω] (3b), and an equation equivalent to the above equation (3) is obtained. Where k ₃
Is usually 1.

In this way, the tolerance α ₁ for the impedance variation and the tolerance α ₂ for the conversion error of the A / D converter are set based on the actual measurement data of the good board, and then the impedance measurement of the test board is started.

If the diode on the substrate is connected in the opposite direction to the above-described state, the reading error will increase during the + half-wave period, so in practice, for each plus-minus half-wave period. Α ₁ (α _1A , α _1B ) and α ₂ (α _2A ,
It is necessary to set α _2B ).

In the above description, the values of the constants k ₁ , k ₂ , k _3, etc. are merely examples, and may be appropriately adjusted according to the actual situation when executing the inspection. The same applies to α ₁ and α ₂ , and if there is no problem, it is desirable to simplify the inspection by rounding the fraction of the value calculated.

It should be noted that FIG. 6 is a flow chart showing an example of the case where the measuring unit 20 automatically sets the tolerance α ₁ for the above variation and the added value α ₂ for the measurement error.

〔effect〕

As described above in detail, according to the present invention, a one-cycle measuring AC signal is applied to each measuring point of the circuit board which is confirmed to be a non-defective product, and the current flowing to each of the positive and negative half waves is added. The corresponding impedance is measured, and the average values Z _SA μ and Z _SB μ are used as the reference impedance at the measurement point, and the root mean square value σ _SA , σ of the average values Z _SA μ, Z _SB μ and the above impedance measurement data. 3 times _SB or 3
σ _SA , 3σ _SB is calculated, and the average value Z _SA μ and Z _SB
Value obtained by adding 2% of μ α _1A = 3σ _SA + Z _SA μ × 2%, α _1B
= 3σ _SB + Z _SB μ × 2% is set as the tolerance for the variation of the test board.

Further, a variable ρ related to the value of the digitally converted data is calculated, and the product α ₂ = Zμ × k ₃ ρ of this variable ρ and a known constant k ₃ and impedance Zμ is associated with the quantization of the A / D converter. As a measurement error, it is automatically added to the above α _1A and α _1B , and is set as the tolerance of the quality judgment for the test board.

Therefore, according to this circuit board inspection method, accurate pass / fail judgment can be automatically performed, including not only the inherent variation in the impedance of the test board but also the measurement error associated with A-D conversion.

[Brief description of drawings]

1 to 6 relate to an embodiment of the present invention, FIG. 1 is a block diagram showing an example of the configuration of a circuit board inspection apparatus to which the present invention is applied, and FIG. 2 is an operation explanatory diagram of each part, Third
The figure is an illustration of the measurement error due to quantization, Figures 4 and 5 are illustrations of the setting principle of the quality judgment tolerances α ₁ and α ₂ , and Figure 6 shows the tolerances α ₁ and α ₂ to the measuring section. FIG. 7 is a block diagram showing a configuration of a conventional device, which is an example of a case where the automatic setting is performed by the automatic setting. In the figure, 11 is a signal source, 3 is a circuit board to be inspected, 13 is a current / voltage converter, 16 and 17 are integrators, 18 and 19 are A / D converters, 20
Is a measuring unit.

Claims (1)

[Claims] 1. A response signal obtained by applying a measuring AC signal from a signal source to a circuit board under test is converted into a voltage and integrated,
The impedance of the circuit board to be inspected is measured by the measuring section using the digital conversion data, and the measured value is compared with a reference value previously taken from a non-defective board to determine whether the measured value is within a predetermined tolerance. In a circuit board inspection method for determining the quality of an inspection circuit board, in regard to the variation in impedance of the circuit board to be inspected, each of the positive half cycle and the negative half cycle of the measuring AC signal is previously subjected to the non-defective product. The first and second tolerances, which are obtained by adding a predetermined value to the 3σ of the impedance distribution characteristic obtained from the substrate, are set, respectively, and the above-mentioned is caused by the quantization error involved in the digital conversion of the integrated voltage of the response signal. For the impedance measurement error of the circuit board under test, Circuit board inspection method characterized by the predetermined corresponding to of come value automatically added to the first and second tolerance the performing quality determination of the circuit board to be inspected.