IPC9850_Surface Mount Equipment Characterization.pdf - 第5页

IPC-9850 Official Proposal May 2001 5 minimizing the measurement-induced portion of the total error observed by the measurement machine, and are particularly important because the specification limits for fine pitch comp…

IPC-9850

Official Proposal

May 2001

inspection systems will improve in their ability to measure component location and orientation. In the future it may

become possible to use in-line post-placement (pre-reflow) automatic optical inspection (AOI) systems to measure

the placement machine capabilities.

Due to the convergence of high-speed and fine pitch machines, this standard makes no attempt to separate the two

types of machines. The user is empowered to decide if a particular model is the best solution for the application

based upon the data reported by the supplier.

1.4 Implementation

1.4.1 Characterization Limitations This standard is comprised of a set of parameters that are the lowest common

denominator for surface mount placement equipment. It is be recognized that additional metrics may be of value in

some instances. The collection of parameters selected for this standard comprise the best subset for use as a core set

of requirements to be included with an equipment supplier’s general specification. This core set may change in

future revisions as technology dictates.

In addition, since there are many possible combinations of hardware and software features that are unique to

individual machine types, this standard cannot address every one of them. Such features and options affect the

overall capabilities of specific equipment model and it is left to the user to understand their implications.

Additionally, it is incumbent upon the user to understand the restrictions and leeway provided for each parameter in

this standard so the proper performance conclusions are reached.

1.4.2 Binding Requirements The body of this document is the standard. The word shall is used throughout this

document whenever a requirement is intended to express a provision that is binding. Material in the appendices is

provided only for information and reference.

1.4.3 Test Components Five component types --QFP-100, QFP-208, BGA-228, 1608C capacitor, and SOIC-16 --

were selected to represent the range of component types placed by surface mount equipment. See Table 3-1 and

section 6 for more precise documentation of these component types.

During the verification of a placement machine, one of the goals is to assess the error induced by the placement

machine. To isolate the contribution of the surface mount equipment to the placement error, which is what this

standard intends to hold suppliers accountable for, it is desirable to reduce other effects that may contribute to the

placement error evaluation process. Using nearly perfect components can best reduce the effect of components on

placement error evaluation. Such components minimize the error associated with SMT component-to-component

variation. For instance, 1608C chip capacitors were chosen as test components because the sides of the capacitors

are very precisely and squarely fabricated. Chip resistors were not selected because they are fabricated in a way that

the top edges of the component seen by the vision CMM may not be in good registration with the side or bottom

features of the component used by the placement machines. The electrodes of ceramic chip capacitors are somewhat

problematic because of their bulbous shape.

SOIC-16 integrated circuits were selected as a standard component because of their relatively low cost and sturdy

construction. SOIC-16 components are believed to be representative of a broad class of coarse pitch leaded

components. Almost all SMT placement machines are capable of processing SOIC-16 devices, so this component

makes it possible to compare the performance of the various machine models. Additionally, 1608C Chip capacitors

and SOIC-16s are economically available in tape packaging, which is how most users feed them because it enables

most machines to maximize the placement speed. Both 1608C and SOIC-16 component types must meet the Joint

Electron Device Engineering Council standard (JEDEC Solid State Technology Association, also known as

JEDEC).

Glass slugs are used to present placement machine vision systems with perfect component images that are free of

bent leads and other part imperfections. The glass slugs also allow the inclusion of fiducial markings that the

coordinate measurement machines (CMMs) use to speed up measurements. The slug fiducial markings are measured

by the CMM, instead of the actual slug component features, to represent the location of the slug component feature.

The slug fiducial markings can only be used to represent the position of the slug component feature if the slug has

been United States National Institute of Standards (NIST) certified to that effect. The slug fiducial markings shall

not be processed by the placement machines during the test and are not known to interfere with the normal operation

of the placement machine's vision systems. The clearly defined slug fiducial markings have the benefit of

IPC-9850

Official Proposal

May 2001

minimizing the measurement-induced portion of the total error observed by the measurement machine, and are

particularly important because the specification limits for fine pitch components (QFP and BGA) are so tight.

For this standard, two QFP and one BGA slugs represent component types. Specifically, the QFP-100, QFP-208,

and BGA-228 component types permit the comparison of an extended range of chip-shooters with that of

specialized IC-placers and multi-functional placement systems. To maintain consistency when utilizing slugs, all

slugs in the pickup trays shall have the same orientation.

Other types of SMT components can be studied using the same basic methods defined in the standard. Glass slug

components that represent micro-BGA and Flip-Chips component types are obvious extensions of the standard that

suppliers or users may wish to consider. Rules for adding claims about additional component types are presented

later in this standard.

1.4.4 Test Panels The evaluation methods of this standard specify component placement into sticky media

dispensed on clear glass panels. This approach is advantageous for two reasons. The first is the dimensional stability

of the glass panel, as opposed to epoxy/glass PWB materials that are much more susceptible to shrinking and

wrapping. The second reason is that it permits use of a standard Optical Coordinate Measurement Machine (CMM)

to illuminate the outline of the components. CMMs are able to rapidly measure a large number of components with

very high precision.

In order to streamline the evaluation methodology, a single glass panel type is specified that permits placement

evaluation and verification of a variety of component types. This panel is referred to in this standard as the

Placement Verification Panel (PVP). The PVP has fiducial markings at specified locations with spacing accuracy

traceable to a NIST certification standard. These fiducials are used as references by both the placement and

measurement machines.

The panels accommodate the following component groups (one component group at a time) for this standard:

a. 36 QFP-100,

b. 30 QFP-208

c. 100 BGA-228

d. 80 SOIC16

e. 400 1608C (Capacitors)

Application of the sticky media requires experience. There must be sufficient amount of sticky media to hold the

component in place, but not too much as to cause interference with the back light intensity. Appendix C has

guidelines on the application of the sticky media.

1.4.5 Measurement This standard has been developed with emphasis on assuring that the measurement tools are

capable of properly characterizing the process. Capable measurement tools provide the customer with a

methodology for verifying vendor claims. They also assure the vendor that the customer is properly evaluating the

machine performance. For an Optical CMM to be a capable gauge for evaluating the surface mount process, its

accuracy and repeatability must be significantly higher that the accuracy and repeatability of the surface mount

equipment. The measurement capabilities required by the CMM depend on the type of components being evaluated

and on the specification limits that the supplier is committing to.

The accepted method for evaluating measurement system performance is called Gauge Repeatability and

Reproducibility (GR&R). This methodology determines the measurement system’s ability to consistently measure

the same product over and over again. The methodology requires that the measurement uncertainty (6x the GR&R

error) be at most 25% of the specification range of the product being measured. GR&R studies do not assess the

accuracy of the measurements, just their consistency. An additional check is needed to verify that the CMM

measurement routine is producing accurate results. See Appendix G for discussion.

The accuracy of the Optical CMM can only be evaluated against a certified gauge. The gauge selected for this

standard is a NIST traceable glass verification panel with etched images in the shapes of the selected components.

This panel is measured on the CMM, and the reported positions are compared to the actual positions of the

component images provided from the panel certification.

IPC-9850

Official Proposal

May 2001

1.4.5.1 Equipment Measurement equipment that meets or exceeds the required GR&R and accuracy specification

limits under the specified evaluation procedures for a specified component type is acceptable.

Automated Optical CMM machines are currently favored by many placement equipment vendors and users due to

their measurement speed. The rapid measurement speed is required due the large number of measurements required

to attain statistically meaningful characterization results. Although this method has been demonstrated to work well,

significant limitations are recognized. The limitations restrict interpretation of the results to the actual SMT process.

Since CMMs were originally developed for a machine shop environment, their lighting systems and measurement

tools are not optimized for measuring surface mount components. The machines are applied to the placement

evaluation process by utilizing clear PVPs that make available high quality images on the CMM. The components

are mounted onto the PVP, are illuminated, and then their positions are measured.

Limitations of the CMMs are their cost and lack of mobility. CMMs are expensive since they utilize many features

that provide structural rigidity, precise position, and high quality imaging. They require delicate calibration and are

sensitive to changes in the environmental conditions, which makes them non-portable.

1.4.5.2 Reporting Measurement results are recorded on standardized forms. The Placement Performance Metric is

form IPC-9850-F1. This form serves two distinct functions. The first is to report the general performance by

machine model type and the second is to validate the performance of a specific machine by serial number. Values

for speed and accuracy that are simultaneously achievable are listed. Form IPC-9850-F1 is supported by the CMM

Capability to Evaluate Metric Parameters form, IPC-9850-F3, which is used to demonstrate the measurement

validity of the parameters presented in the Placement Performance Metric.

The Reliability Performance form (IPC-9850-F2) provides an indication of typical results determined from values

reported by users. Unlike the Placement Performance Metric, which cites guaranteed performance, the Reliability

Performance form merely conveys typical reliability levels for the machine model type based upon a nominal

amount of data collected by the placement machine supplier’s customers in their factories. But this reliability

performance form does establish a foundation upon which the industry can raise the reliability, availability and

maintainability of the surface mount equipment infrastructure.

1.5 Forms Reproducible copies of all required forms are provided in Section 7.

1.6 Data Methods

With regards to the statistical treatment of placement deviation data, we have assumed that deviations generally

follow normal (Gaussian) distributions. This may not always be the case. There is nothing inherently wrong with

data having some other distribution. However, depending upon the circumstances, it might mean that the

specification limits at the two prescribed Cpk limits (1.33 and 2.0) contain some percentage of data less than the

99.9968% and 99.9999% percentages normally associated with these Cpk limits. It would be prudent for the user to

verify that the machine has data that reasonably approximates a normal distribution if it is crucial to the user that the

machine's level of inaccurate placements precisely matches those associated with Cpks of 1.33 or 2.0.

1.7 Terms and Definitions

Additional terms and definitions are provided in Table 4-4 and 4.1.1 through 4.1.11.

Fiducials are markings on substrate materials --such as PWBs or glass-- that are utilized by the placement

equipment vision systems to locate the position and orientation of the substrate.

Component refers to surface mount parts which are utilizes in the construction of electronic devices.

Placement Error is defined as the physical distance between where the component was actually placed and where

the surface mount equipment was specified to place the component by the placement program data.

Slug refers to a NIST certified glass panel with etched image of a single Component and reference fiducials.

dev

refers to the placement error in the X-direction (parallel to the “9850 Verification” label found on the

Placement Verification Panel).