Search results

The availability of a wetting balance which can be easily interfaced to a microcomputer has made possible a practical receiving inspection solderability test for component leads that avoids the subjectivity of the present dip‐and‐look test. The wetting balance, in effect, detects the size and shape of the solder meniscus on the lead. Since it is the solder meniscus more than the degree of coverage that is evaluated by inspectors of the completed solder joint, the wetting balance provides a more realistic test of how well the components will perform on the PWA. The software that has been developed for the wetting balance is designed to make it easy for inspection workers to perform the test with a minimum of training. It asks for identification of the part, manufacturer, date code, purchase order number, etc., so that the final results are adequately documented. Use of a computer to present the results means that the wetting force as a function of time can be plotted as a normalised curve (automatically accounting for differences in number and size of leads), and also that the results can be accumulated in a factory computer for statistical quality control. For a given lot of components, there is usually little spread in the observed results. This indicates that a sample size as small as three is sufficient to characterise the lot. With further data accumulation, it should be possible to devise a skip‐lot sampling plan for those manufacturers showing consistently good solderability. Also, the accumulated results of lots from problem manufacturers, coupled with microscopic studies of the causes of poor solderability, can be used as a basis for negotiations. The results are reproducible from one plant to another because they do not require visual interpretation. Judicious application of this method of solderability testing by a component user should allow removal of problem lots (for return or solderability enhancement), and therefore lead to a virtual elimination of solderability‐related defects observed after PWA soldering. Widespread application by component manufacturers (after burn‐in) should lead to a virtual elimination of cases of shipping unsolderable components.

This paper discusses some basic ideas about process development and control in Part I and applies them to soldering in Part II. Because it is possible to understand how design, materials and process affect the product, it is unnecessary and inappropriate to resort to the statistical‐correlation methods that are applied to complex processes. A process qualifies for the label ‘closed loop’ only if the design and materials going into.it are controlled. The types, degree and sophistication of control needed for a process are to be judged by consistency of the product. For soldered assemblies, the product is evaluated by visual inspection, and the adequacy of process development and control depends on the adequacy of inspection. Inspection can be improved if it is regarded as a process. It can also be improved if inspectors understand which features are important and which can be ignored safely, i.e., by understanding their causes and associated risks. Much of the criticism of visual inspection, and perception of need for automated inspection, derive from a failure to distinguish clearly enough between material and process variables, between the two types of inspection (product‐oriented and materials/process‐oriented) and between appearance and risk. Properly controlled visual inspection is well suited for evaluating the soldering process. The most important visual attribute to look for in solder inspection is the contour of the fillet, because this is what reveals the quality of wetting, and wetting is the most important physical attribute of the connection in determining its strength and reliability. Wetting depends on just two basic requirements, heat transfer and solderability, and these are discussed in some detail. Causes of non‐ideal texture and lustre of the solder are given, but these attributes do not affect reliability, nor is measuring solder purity important. Additional factors which do affect reliability relate more to design and materials than to process. Failure to deal with these factors can result in solder defects that are undetectable by any inspection technique. The answer to this problem is therefore not automated inspection to find more kinds of defects than visual inspection can, but control of design and materials, as well as process, to prevent them entirely.

A challenge in selecting and applying lead‐free solders lies in separating the influences of materials' properties, fluxes and processes to obtain robust assembly conditions that are compatible with PCB finishes and all component terminations. This paper discusses simple steps towards establishing a lead‐free assembly process. With reference to results of solder paste spread and wetting tests and component solderability tests, some of the current limitations in applying standard test methods to lead‐free evaluations are highlighted.

The purpose of this paper is to investigate the effects of H2 flow rate on improving the solder wettability of oxidized‐copper with liquid lead‐free solder (96.5Sn‐3Ag‐0.5Cu) by Ar‐H2 plasmas. The aim was to improve the solder wettability of oxidized copper from 0 per cent wetting of copper oxidized in air at 260^oC for 1 hour to 100 per cent wetting of oxidized‐copper modified by Ar‐H2 plasmas at certain H2 flow rates and to find correlations between the surface characteristics of copper and the solder wettability with liquid lead‐free solder.

To reduce the copper oxides on the surfaces of oxidized‐copper for improving solder wettability with liquid lead‐free solder, this study attempted to apply Ar‐H2 plasmas to ablate the copper oxides from the surfaces of oxidized‐copper by the physical bombardment of the Ar plasmas and to reduce the surfaces of oxidized‐copper by the chemical reaction of H2 plasmas with the surfaces of oxidized‐copper.

The solder wettability of oxidized‐copper was found to be highly dependent on the surface characteristics of the copper. The values of polar surface free energy and dispersive surface free energy on the surfaces of oxidized‐copper modified by Ar‐H2 plasmas were close to those values of solid lead‐free solder, which resulted in improved solder wettability with liquid lead‐free solder. Auger spectra indicated that the Ar‐H2 plasma modification was used to remove the copper oxides from the surfaces of oxidized‐copper.

The surface characterization of copper surfaces is typically determined by expensive surface analysis tool such as Auger Electron Spectroscopy (AES). This paper reports the results of a study of a promising technique called the sessile drop test method, for examining the surface free energies such as total surface free energy, polar surface free energy and dispersive surface free energy on the surfaces of copper to clarify how the solder wettability of oxidized‐copper with liquid lead‐free solder was enhanced by Ar‐H2 plasmas.

Rapid feedback on the effect of changes in materials or assembly process parameters on solder joint quality is essential in process optimisation for fine pitch surface mount solder assembly. This is particularly the case where the demands of high volume production do not allow lengthy experimentation to be carried out on the production line. The greater precision and care required in analysis of fine pitch solder joints places a further constraint on the speed at which the analysis can be carried out. The techniques of microstructural analysis, solder joint mechanical strength testing and short duration environmental stress testing must be used to meet a rapid turnaround quality and reliability analysis requirement and to maximise the amount of information obtained from each stage of the analysis. Three case studies are detailed which demonstrate the use of these techniques in a high volume production context to provide rapid feedback on joint quality. The case studies have been selected for presentation not only to demonstrate the analysis techniques but also because they address issues which are of current interest due to the increased usage of fine pitch packages in production and the constraints on the use of solvents for cleaning of circuit boards. The studies are:

An investigation has been performed on the improved solder wettability of oxidized aluminum (Al) with lead-free solder (96.5Sn-3.5Ag) using Ar-H₂ plasmas. The lead-free solder wettability was raised from 62.2 per cent wetting for Al oxidized in air at 250 C for 4 h to 98.4 per cent wetting of oxidized Al modified by Ar-H₂ plasmas at a certain H₂ flow rate. This study aims to gain insight on the surface characteristics of Al affecting the solder wettability with a liquid lead-free solder.

Ar-H₂ plasmas at certain H₂ flow rates are intended to reduce Al oxides on the surfaces of oxidized Al substrates both by physical bombardments via Ar plasmas and chemical reductions with H₂ plasmas, while Al substrates are exposed in Ar-H₂ plasmas to improve the solder wettability with a liquid lead-free solder.

Surface characteristics of oxidized Al substrates have been identified to play key roles for enhanced lead-free solder wettability using Ar-H₂ plasmas. A decrease in polar surface free energy and an increase in dispersive surface free energy on the surfaces of oxidized Al substrates are exploited to advance the lead-free solder wettability. Decreased composition ratios of O to Al, detected by X-ray photoelectron spectroscopy (XPS) for oxidized Al substrates, are crucial for improved lead-free solder wettability.

XPS is typically used to analyze the surface compositions of Al oxides. To provide a rapid and non-expansive method to identify the surfaces of Al substrates prior to soldering to assure lead-free solder wettability, this study proposes a measurable skill, a so-called sessile drop test method, to investigate surface free energies such as total, polar and dispersive surface free energy on the surfaces of Al substrates, to illuminate how the lead-free solder wettability of oxidized Al is improved by Ar-H₂ plasmas.

The use of nitrogen‐based soldering environments, rather than ambient air, has been widely promoted for both reflow and machine soldering. Earlier claims of process improvements derived mainly from the enhancement of process windows related to improved solderability. The improvements seem to be accepted in the industry but may not be very significant except when accompanying other changes in production soldering facilitated by the use of inert environments. Appropriate process changes include the use of finer pitch devices and more densely populated boards and changes in cleaning technology or the use of no‐clean processes. Material changes include the increased use of new technology for both pastes and fluxes (generally no‐clean), increased use of high temperature solder alloys and the possible need to use lead‐free alloys for environmental reasons. This paper discusses the relationship between atmosphere effects on soldering processes, and process and material changes which favour soldering in atmospheres other than air. It presents some recent data on improvements to soldering resulting from the use of nitrogen environments.

During reflow soldering the applied solder paste is melted and the components, previously placed on the solder paste, move into their final position. This process, however, may be accompanied by various unwanted movements of components and solder. Components may move horizontally along the surface of the board (this is called swimming or floating), or may move vertically and stand on their ends (this is called drawbridging or Manhattan effect). On the other hand, the molten solder may move to places other than those intended, e.g., into metallised holes (PTH) connected to the solder lands, or upwards along component leads away from the joint area; this effect is called solder wicking. Moreover, isolated small solder balls are often found on the board surface after melting of the paste. Experiments show that all these effects depend on the heating method, vapour phase soldering often being the most prone. The driving forces of the displacements can be explained in terms of forces and pressure caused by the surface tension of the molten solder, whereas the observed influences of the heating method are the result of the direction from which the heat is transported to the solder paste to be melted. From this, important conclusions for vapour phase soldering, infra‐red soldering and hot‐belt soldering may be drawn.

The reduction in IC package lead pitches in surface mount solder assembly and the current high emphasis on quality and reliability of printed circuit assemblies have created a requirement for microanalysis of fine pitch solder joints in manufacturing situations. Of particular interest are metallographic analysis, detection of solder joint defects and mechanical strength testing of solder joints. Much has been published in the literature on the results of such evaluations in specific applications but little has been available on procedures for use in the microanalysis itself, particularly for fine pitch solder joints. Detailed procedures for fine pitch solder joint microanalysis, which the authors have verified down to 0.5 mm (0.02 in.) lead pitches, are presented. In particular, the authors present procedures for metallographic examination of tin‐lead and tin‐lead‐silver solder joints. In addition, test parameters are given for a repeatable technique of fine pitch solder joint mechanical strength testing that allows mechanical strength measurements to be obtained from almost every lead on a fine pitch surface mount IC package.

Results for reflow soldering are presented from a three‐year EC funded project “IDEALS” to develop lead‐free soldering solutions. On the basis of fundamental data from the literature, a shortlist of candidate lead‐free solders was selected, and results from tests on physical and soldering characteristics, and wetting balance testing, led to the choice of SnAg3.8Cu0.7, melting at 217°C. Implications for solder paste medium development are discussed. Differences in alloy density, melting point, and surface tension relative to conventional solders were found to give higher levels of internal voids, reduced spread on copper, and rougher, duller joints. Reflow process window studies showed that sound reliable joints could be obtained with a peak temperature as low as 225°C. Reliability was tested on soldered test boards using thermal shock cycling, power cycling, and vibration. Overall the SnAg3.8Cu0.7 gave results approximately equivalent to conventional solders, and different board finishes had no significant effect. The effects of Sb and Bi were also evaluated. No justification was found for minor additions of Sb, but 2‐5 per cent Bi was found to allow a reduction of the peak reflow temperature, though at the cost of reduced reliability if any Pb was present.