SMART Events

The weather was extremely bad and consequently several delegates were delayed but a full attendance was finally achieved, demonstrating the interest in the Irish electronics industry to understand the lead-free legislation and implementing a lead-free process.
Forty plus delegates were welcomed by Philip O'Rourke Chairman SMART Group Ireland and Vincent Synnott SMART Group committee member. Seven invited speakers were supported by desktop exhibitions promoting lead-free solders, rework tools / equipment and manufacturing commodity supplies.

Speakers Left to right Kieran McGee, SMART Group - Alan Brewin, NPL - Brian Allder, Qualitek - Richard Boyle, Henkel - Angus Westwater, Rohm - Philip O'Rourke, Chairman, SMART Group

The following is a brief resume of each presentation. It was comforting to see that all the research and lead-free trials being reported have a common message and confirm lead-free soldering is possible and increased joint reliability will be achieved. There was no contradiction within the presentations, which were prepared in isolation. It is clear that leadfree technology research has culminated down to SnAgCu, SnCu and Sn technologies for component plating, paste and bulk solders and reliability testing has given a green flag at least until long term experience may contradict this.

The WEEE and RoHS directives drive the core of the lead-free legislation. Gordon Pryor STC Optronics supported by Nick Jolly DTI UK, presented the past history of the WEEE directive leading up to the formal acceptance of the directive by the EU on the 13th February 2003, to come into force on 1st July 2006. It was made clear that July 2006 is the date that products cannot be sold containing lead, except for specifically temporarily exempt applications. Exemptions will be reviewed on and around the 13th February 2005. Within the scope of the WEEE and RoHS directives, Lead, Mercury, Hexavalent Chromium, Cadmium, and flame-retardants, Polybrominated Biphenyls (PBB) Polybrominated Diphenyl Ethers (PBDE) are banned. The definitions of "Producer" "Put on the market" and maximum allowable concentration of banned substances were explained and expanded but many questions from the audience were asked concerning policing the directive. Several areas of who is responsible for what remained unanswered awaiting the results of ongoing consultation between the member states. Nick recommended regular viewings of the DTI web sites to read the current status on many activities ongoing at this time.

With respect to future obligations by producers, users and recyclers, the directive draws a financing distinction between product put on the market before 13th August 2005 and product after this date. The strong message was the EC's RoHS and WEEE directives are a reality at a European level and individual member states are introducing collection and recycling goals during 2004 with the legislation taking effect during 2005 onward to full implementation 1st July 2006.

Gordon also gave details of the "LEADOUT" project, which is a 5 million Euro project funded by organisations in Portugal, UK, Spain and Hungary. The project starting early in 2004, will assist small and medium sized companies in Europe and emerging countries to make the change to lead-free soldering

Finbarr Waldron, NMRC Ireland, gave a comprehensive presentation of product design and process optimisation philosophy to maximize reliability when making the transition to a lead-free process. He also discussed in detail the thermal, mechanical and environmental factors, which affect product reliability. Fundamentally he was presenting "Get it right first time".

Comparisons of what made Sn63Pb so good against proposed lead-free alternatives were made and best alloys to replace SnPb were discussed and their pros and cons. Several very interesting examples of component and solder joint failures via modeling techniques were highlighted and how they could be avoided.
Finally, design for reliability theme was concluded with reliability improvement techniques including FMCA, FTA and FRACAS (tolerance, SPC and redundancy).

Angus Westwater, Rohm Electronics, presented the fundamental reasons why lead-free solders were realised and how lead and its combination with tin has played a major roll in our industry as a low cost very well understood high reliability alloy. The current leaded soldering process was presented to illustrate where SnPb alloys are positioned in the product assembly and latter usage thermal spectrums, and the relatively comfortable process windows that are currently enjoyed.

The philosophy to introduce lead-free termination components into existing leaded process and make no change to process parameters was strongly promoted via reliability data, and several examples of joint failure modes via lead contamination were given, where leaded terminations were used with lead-free solder assemblies. The qualification of components to the higher soldering temperatures was raised and again the dangers to use leaded components in the increased temperature environment of a lead-free process.
Component solder plating via electroplating and solder dip were explained and the definition of solderability reviewed. Several graphs showing maintained mechanical reliability with lead-free in leaded and total lead-free process was given and examples why lead-free process temperatures must be increased to achieve solder wetting times below 2 seconds.

Using pure tin (Sn) as component plating was defended with respect to tin whiskers and the dramatic reduction in the probability of whisker growth after the plating had been alloyed with the base solder.

Finally, delegates were encouraged via examples, to immediately implement a transition to lead-free soldering program via a formal road map, and work to fully understand their supplier base elimination of lead programs.

Solder specialist Brian Allder, Qualitek-Europe, reminded the audience of the original reasons to remove lead from our environment via examples of how lead toxicity affects humans and our fellow creatures. Some detail was presented on several lead-free alternatives, their cost and assembly process implications. With the slower wetting characteristics of lead-free solders and the proposed use of more active water-soluble fluxes, Brian discussed dross, flux chemistry and management, and the possibility of introducing board washing again.

Lead-free soldering equipment within wave soldering processes was reviewed and recommendations given to introduce a program of equipment upgrading to compensate for the high tin content solders and bath erosion. From experience with customers working with lead-free solders and the knowledge that lead-free joints are prone to voiding, Brian presented evidence that baking the bare PCB at 110ºC for two hours dramatically reduced the occurrence of voiding, showing that the cause of voiding is contributed by humidity ingression in the board exaggerated by the increased lead-free soldering temperatures.

Phil Atkinson, Cookson Electronic Equipment, gave a presentation very much based on the practical issues and costs incurred when making the change to a lead-free process. Comparisons of several lead-free alloys were made and their increased cost comparisons against SnPb. Main impacts on existing processes is the 20 to 40C increase in soldering temperature and the narrower process window. Delegates were made aware of patent issues that exist with some lead-free alloys.

Covering the existing process parameters, Phil highlighted that the stencil printing process is probably the one process that is not significantly affected by the change to lead-free soldering and that existing stencils, squeegees, wiper papers, solvents and print parameters all remain the same. To achieve and maintain higher soldering temperatures within the smaller process window dictates that convection heating is required with the use of nitrogen achieving improved pad wetting demonstrated via several excellent video clips.

Much time was spent on air versus nitrogen flow with the benefits of nitrogen offset against running costs, equipment purchase and the need for equipment modification to introduce nitrogen gas into machines incorporating cooling zones, this lead onto much discussion, on flux management. Strong message was that lead-free alloys have a significant impact on wave soldering. Bath solder temperatures need to be set between 260 - 270C and solder drainage will be poor with nitrogen atmospheres being a real advantage. Aqueous cleaning should be considered where increased flux activity is introduced.

Alan Brewin, National Physical Laboratories (NPL), presentation focused on the science of reliability applied to solder joints, laminate insulation (Electromigration / Dendrites) and component loss of functionality, sharing much information from NPL testing programs.

Thermal cycling and mechanical fatigue reliability failures were demonstrated and measured against mismatch of TCE between component and PCB materials. Excellent data relating to modeling of joint failures during stress and strain cycling for large and small components was shown, with an interesting example of BGA termination failures being restricted to around the edge of the die. This being the area of maximum differential of TCE between the die, 2.8 ppm/ºC and the PCB, 18ppm/ºC. Interesting conclusion in this section was all alloys respond differently under differing strain ranges. SnPb can accommodate more strain, but SAC will be more reliable in low strain cycling environments, which is the majority of applications today. Using video clips, Alan demonstrated the more brittle nature of lead-free terminations against the more ductile SnPb alloy although the tensile stress figures are similar. Lead contamination of leadfree joints does reduce the joint thermal and mechanical fatigue capability but increasing the contamination to 5% and above, was shown to increase the durability of the Pb free joint.
Component and PCB dimensional stability (TCE) is a concern with increased soldering temperatures. Alan presented data and images showing greater than 100 microns variance immediately after reflow between QFP terminations and PCB pads. With 0.3mm pitch high pin count QFP's; this expansion differential will not be acceptable.

The phenomena referred to as conductive anodic filamentation (CAF), is seen within the weave of epoxy glass FR4 PCB substrates. The result is a reduction in insulation resistance in multilayer boards between vias and power or ground planes. Images were presented showing this phenomena and time to failure graphs which were worryyingly short (approx 100 hours). Models were presented showing attention to weave orientation can dramatically increase the time to failure. Humidity is a major driver in CAF and with an increase in laminate failures directly attributed to lead-free soldering temperatures.

Richard Boyle, Henkel. Running a lead-free process throws up several issues associated with high tin content solders, excluding these directly associated with the actual solder joint. With leaded solders any tools or fixtures dropped into the solder pot will float on the surface making retrieval easy. With the less dense lead-free solders, tools etc will sink adding to possible solder contamination and causing potential damage to the machine. Operators must be educated to ensure tools are connected to lines to ensure retrieval (not tied to the operator). From previous presentations during the day, Richard confirmed the grey appearance of lead-free joints against the shiny SnPb materials and the need again for inspection personnel retraining.

The higher surface tension of lead-free solders and hence wetting of PCB pads during reflow produces the effect of a copper hallow on each joint. Richard showed how nitrogen gas introduction to the system dramatically improved wetting and joint appearance. Copper build up in wave soldering systems causes intermetallic particles and solder bridging defects. The copper disillusionment in the bath can be corrected with the addition of Sn3.6Ag at approximately 3 to 7 month intervals depending on the processes and through put. Also, nickel accumulation in the bath is the result dissolution of nozzle and pot materials. Nickel increase is observed after 2 to 3 months and can be halted with the addition of fresh solder.

Lead contamination of the solder bath will remain a problem for the future until all leaded component terminations have been eliminated. It is a particular concern where both bismuth and lead are contaminates, inadvertently placed in the same bath, as melting point phases as low as 96ºC can be generated, producing low reliability joints and dewetting and reduced creep parameters during product life. Richard again highlighted the concerns of fillet lifting a phenomena associated with wave soldering of through-hole devices. The bottom line was fillet lifting although not cosmetically desirable does not reduce the joint mechanical reliability.

Report by Angus Westwater, SMART Group Committee

Special thanks to Angus for putting pen to paper on the way back from the event.

Bob Willis SMART Group Web Site Coordinator

Back to events