IPC9850_Surface Mount Equipment Characterization.pdf - 第25页

IPC-9850 Official Proposal May 2001 25 Suppliers may include procedures to measure placements with the placement machine itself. This is acceptable for a particular placement machine as long as that particular machine (n…

IPC-9850

Official Proposal

May 2001

4.2.9 Preventative Maintenance (PM) Time

This is the amount of time required for preventative maintenance as specified by the machine supplier. This metric

will be based on the maintenance routines that are required each year in order to keep the machine in a warranted

condition and how long those routines take for a 6000 hour period.

Calculation Method Multiply the amount of time required for each PM procedure by the number of times it is

required in 6000 hours and add these up for a grand total.

4.2.10 Mispick Rate (in PPM) This is the frequency with which a component was not picked or picked incorrectly,

even if recovery handled the problem. Mispicks include both pickup failures and vision failures, but excludes

components exhaust.

Pickup defects will be handled in a manner similar to placement defects. Pickup defects are categorized as vacuum

failures or vision failures. While many suppliers have various auto-recovery routines that prevent a pickup defect

from resulting in a missing component placement defect, this type of defect has still been deemed worthy of tracking

and reporting due to its impact on the assembly process. Pickup defects cause a reduction in net CPH and the

unnecessary rejection and scrapping of good parts.

The number of mispicks that occur during the 4400 part placements made to determine placement defect rate will be

recorded on Form IPC-9850-F2. The number of pickup defects made during this process for the past 20 machines

will be reported, from which a point estimate for pickup defects in PPM can be made.

Measurement Procedure Clear the Mispick counters in the machine. Empty the reject bin. Record the relative

humidity. Run the attribute defect test (place 4,400 components). Record the vacuum failure, and vision failure

counter values. Examine the components found in the reject bin and count the defective components.

Calculation Method The total pickup attempts minus total number of placements minus the number of defective

components divided by total pickup attempts minus defective components, multiplied by one million to obtain ppm.

Mispick Rate =1,000,000 * (Total Pickup Attempts)- (Total Number of Placements) –(Number of Defective Components)

(Total Pickup Attempts) –(Number of Defective Components)

4.3 Reporting Sites This is the number of machine sites included in the reliability metric calculations.

4.3.1 Number of Machines This is the number of individual machines included in the reliability metric calculations.

4.3.2 Total Number of Placements This is the number of individual machine placements included in the reliability

metric calculations.

4.3.3 Total Pickup Attempts This is the number of times the nozzles attempted to pick components.

5 Measurement Capability Verification –Form IPC-9850-F3

Form IPC-9850-F3 functions as a tool for demonstrating the measurement capability of the measurement equipment

used to generate the parameter values for the Placement Performance Metric (form IPC-9850-F1). The measurement

capability information presented on the form shall be gathered and reported according to this standard, and shall

reflect the performance of the measurement equipment used in the machine performance analysis. The reported

information should be acquired no more than 90 days prior to the machine performance study.

The parameters of this metric shall evaluate the measurement system’s ability to accurately measure the test

vehicles produced during the machine performance evaluation analysis in a repeatable and reproducible manner.

This evaluation shall be made for each component type and yields results about the measurement capabilities of the

CMM along the same axes as reported by the placement performance metric. exceed The specification limits

reported in the performance reporting form shall never be smaller than the larger of the GR&R and accuracy CMM

specification limits per component type. Suppliers that add additional component types to the Placement

Performance Metric shall also report their measurement capabilities for the same component types.

IPC-9850

Official Proposal

May 2001

Suppliers may include procedures to measure placements with the placement machine itself. This is acceptable for a

particular placement machine as long as that particular machine (not just a sample machine from that particular

model type) passes the gauge requirements laid out in this section which determine whether or not its measurement

capability is acceptable.

5.1 Gauge Repeatability and Reproducibility Capability

This verification procedure determines the CMM's ability to repeat and reproduce the component location and

rotation mounted onto the PVP. This evaluation shall be made for each component type.

While multiple GR&R methodologies are available, this standard uses the traditional average-range method and

recommends the use of the spreadsheet discussed in Appendix H. The results are reported in terms of the upper and

lower specification limits that yield a precision-to-tolerance ratio of better than 25%.

Although a precision-to-tolerance (P/T) ratio of less than 20% is usually considered acceptable, 25% is utilized in

this standard to accommodate existing CMM machines. However this reduced P/T ratio specification if somewhat

offset by the fact that the analysis is performed using a more stringent requirement of six sigma (+/-3 sigma or

99.7%) versus the customary value of 5.15 sigma (+/- 2.575 sigma or 99%) recommended by the Automotive

Industry Action Group (AIAG).

Measurement Procedure Obtain a single glass verification panel with a specific part type populated according to the

procedures described in the machine placement performance metric, Section 3. Measure the populated PVP three

consecutive times. The PVP is to be completely removed and replaced from the CMM for each run. Only 36

components need to be measured for each run. Appendix F outlines the component locations for the specific

component panel layout. After the three measurements are obtained the CMM must be shutdown and restarted.

The same PVP plate shall be subjected to two additional measurement cycles (two more sets of 3 measurements)

using an identical procedure. At least two operators should be utilized to obtain the three sets of data. The entire

GR&R evaluation shall be performed in a time period no longer than it takes to perform the measurement of the

consecutive four PVPs for the placement performance evaluation.

For the SOIC-16 and 1608C components, the center of the component shall be measured using the lead outline and

component outline, respectively. The same measurement procedure applies to the PVP populated with slugs.

However, since the CMM utilizes the fiducial markings on all glass slugs, regardless of the specific component

image, it is only necessary to evaluate the CMM’s ability to measure the slugs using the fiducial markings and not

individually for the QFP and BGA component types. Appendix D provides suggested methods for measurement

methods.

5.2 Accuracy Capability

This verification procedure determines the Optical CMM's ability to accurately measure the placement of

components on the glass verification panel. This is accomplished by evaluating the CMM’s ability to measure the

NIST certified glass verification panel specified in Figure 5-1. This accuracy verification panel (AVP) was designed

to evaluate the CMM’s ability to measure three component types required for compliance with Section 3.

To evaluate the measurement capabilities of the 1608C and SOIC-16 components, their images have been etched

onto the AVP. Since slugs are utilized for the QFP and BGA components, only slug fiducials marking are etched

onto the AVP. Some of the etched images are translated and rotated from their specified CAD location, in order to

test the CMM’s ability to measure components with measurement error. Some of the 1608C and SOIC-16

component images on the AVP are translated in a range of + 50 micrometers, and rotated in a range of +3º range.

Some of the four fiducial image sets will be translated in a range of + 25 micrometers, and rotated in a range of +1º

range. The ‘CMM Accuracy Eval.xls’ file provides an instruction guide and a sample CMM accuracy evaluation

spreadsheet.

This methodology is implemented using a measurement program on the CMM, which searches for the components

on the glass panel at their CAD locations. These CAD locations are the centers of the components as if no

translation or rotations were applied to the images on the glass panel. The CAD locations of the images can be found

in Appendix F. The CMM measurement results are compared to the locations and rotations of the component images

IPC-9850

Official Proposal

May 2001

provided by the NIST certification for the unique glass panel utilized. The closer the CMM measurement values are

to the certified locations the better the measurement capability of the CMM.

The results are reported in terms of the upper and lower specification limits that yield process capability indexes

equal to 2.0. See Appendix B for discussion.

Measurement Procedure Measure the center location and orientation of the component images on the certified

accuracy verification panel a single time. Measure the center location and orientation of each group of four fiducials

that represent the slug component types. Use the same image analysis procedures that are utilized for the Placement

Performance Metric evaluation. Appendix D provides a guide for the proper methods of measuring the center

locations of the components.

Calculation Method Subtract the CMM measured image locations from the NIST certified image locations along the

X, Y and θ axes, for each of the 1608C, SOIC-16, and slug features. This provides a distribution of the measurement

errors of the CMM. Use these distributions to calculate the average and standard deviation for each axis. Calculate

the process specification limits for each axis using the following equations, when Cpk = 2.0. See Appendix B for a

discussion of these equations.

SL=3*SD

Error

*Cpk+|Avg

Error

where Avg

Error

∑

−=

)(

ocationCertifiedLcationMeasuredLo

and SD

Error

)(

118

ocationCertifiedLcationMeasuredLo −

−

∑

Figure 5-1 Accuracy Verification Panel