SMART GROUP 26th Annual Conference, The Oxfordshire Golf Club, Thame, Oxfordshire, October 6 and 7 2010
Article Type: Conferences and exhibitions From: Soldering & Surface Mount Technology, Volume 23, Issue 1
October 6
Keith Bryant, SMART Group Chairman, cheerfully bade all the delegates a warm welcome to their annual conference, held once again in the elegant surroundings of the Oxfordshire Golf Club (Figure 2).
The first event of the day was an interview of Robert Gregory of Rolls-Royce by Graham Naisbitt, SMART Group Vice-Chairman. Robert Gregory is on the International Advisory Committee to Pb-Free Electronic Risk Management (PERM),studying lead-free electronics reliability.
One of the major issues facing industry, said Bob Gregory, was high-reliability design. The supply chain has broken down was his contention. His particular field at Rolls-Royce is with nuclear submarines, and here it a critical matter of knowing why things fail. Components are designed for a 40-year lifetime, have to withstand multiple shocks, and he needs to know who is designing what with what. People and materials evolve, but manufacture is now done elsewhere, which means that the quality of materials he receives is not always fit for purpose, notably laminates, and so one becomes a little sceptical. Testing on all components has to be done, that is 100 per cent testing, ensuring that all calibration tests are done, but it is now apparent that the supply chain is out of control with regard to quality, and that is a major concern. One link in the chain is the use of counterfeit components, and there was an instance where a registered and franchised distributor supplied counterfeit components, with a certificate of conformity, which is a real worry.
Graham Naisbitt asked if he had any misgivings about the refinishing of components. Bob said that he did, especially when there is a risk of tin whiskers, such as is the case when tin is coated with tin-lead, for example. Were there problems with lead-free soldering? asked Graham. Yes, the problems are with the temperatures required, which bring about other problems, such as fluxes. Lead-free means tin whiskers, so all surfaces will have tin whiskers over their operational life, and plenty of them, too, and they will be long enough to give problems.
Prompted by a question about consumer electronics having built-in obsolesce,which increases the volume of waste material going to landfill, Bob had some fairly astringent views on the WEEE directive. Lead in tin-lead is inert, and is safe. Silver leaches out, lead does not. He believes that most people do not know what they are doing and are being advised by people who are biased, which is dangerous, generally. We live in age of science where reason and sense have flown out of the window, he concluded, and lead-free solder will never be suitable for a military application, certainly, none of the present lead-free solders are fit for purpose in the military sense.
Dr Chris Hunt from NPL spoke to the meeting about reliability issues with lead-free solder, which was brave considering the previous conversations. Chris is also working on the $100 million PERM project which runs organised research activities. One involves SAC solder balls which are more rigid, and there is an interfacial fracture tendency. Some of the issues include repair,aging phenomena, shock/vibration testing, examination for interfacial fractures,etc. It has been noted that SAC shear strength will drop to zero after seven years of exposure, which, one imagines, makes design for a 40-year life rather problematic.
Solder fatigue: the work being done at NPL on this included testing solders to failure point. They have a purpose-built test jig, which can measure properties such as creep, fatigue and relaxation. With electromigration effects,NPL are studying the combined application of stress, temperature and high current density. They also have a vibration properties testing instrument, which is designed to measure solder joint reliability up to 1 kHz.
He covered two further problem areas, one being tin pest – in which transformation rates are measured for tin alloys to degrade. This was shown in a video, and tin pest is self-seeding, not perhaps a good thing to happen in a circuit board in, say, an aircraft. Some fluxes appear to arrest the transformation. The other area of work was on CAF (C for cathode, A for anode,and F for failure) on PCBs, a project in which there are 11 partners including Graphic plc. Finally, he came to coatings – mitigation and adhesion– Chris showed a parallel-plate test vehicle, in which copper coupons are strung along an insulated rod with spaces of suitable intervals, and whisker growth can be measured. Adhesion was measured with a peel tester. There is a great deal of work going on with lead-free solders, but one is bound to conclude that none of this would be necessary if the green lobby had been better informed some years ago. The cost of ignorance, a subject dear to the heart of our next speaker.
Doug Pauls from Rockwell Collins sees life through experienced and perceptive eyes, and gave, as usual, an amusingly delivered paper, this one on the impact of handling PCBs and the correlation of cleanliness and SIR results. He went back to one winter when Cedar Rapids, the home of Rockwell Collins, suffered from major flooding which disrupted the water supply. No water meant no washing,so how to clean the hands of those employed on the manufacturing floor? Hand sanitizers and hand lotions were installed in the work place. Now bear in mind that Rockwell Collins are fanatical about cleanliness; they do not require gloves to handle assemblies as there is a final aqueous cleaning step in place prior to conformal coating. The ones they selected were ethanol-based sanitizers and lotions too. Doug’s department undertook a study of these products, and through chemical analysis of sanitizers and lotions, concluded that they were all pretty similar in form. They applied the lotions, etc. to alumina material as the substrate and looked for residues. Ion chromatography results were interesting; there was some sodium transfer but this was infinitesimal. Ionic material transfer was no worse than with bare hands, and by testing over a period of time, at temperature and humidity, it was shown that there was no corrosion or metal migration, and no signs of coating degradation. Happily,their water supply is now fully restored and river levels are healthy, but it was good to know that some of the exotically named lotions employed during that fragrant period had no adverse impact on production.
David Hillman is the other half of the team from Rockwell Collins, but he is a serious metallurgist and thus able to speak knowledgeably about voiding in BGAs – to void or no to void. Hamlet would have appreciated this. Yorick had a lot of voids. There are many void types and causes; void types include macro voids (also known as process voids, as they stem from the use of volatile fluxes); planer voids, which have a root cause in an incorrect-plating process(Intel have had a long look at planar micro voids, and have concluded that the effect on solder joint integrity was due to temperature shock); shrinkage voids,also known as hot tear voids (the main cause is solidification of a solder alloy) and finally micro-via voids, which have their root cause in volatile flux content. Of these, a pinhole void can be traced back to the plating process and a Kirkendall void has solder diffusion as a root cause. Do voids affect solder joint reliability? There are arguments both for and against, but generally speaking, yes. However, voiding accounts or less than 10 per cent of the solder joint area. The ones to have, it seems, are champagne voids, which shows a bit of class.
RoHS and ReacH – are they killing the industry in Europe? was the pertinent question posed by Nigel Burtt of Ennjaybee Associates. The word“yes” immediately comes to mind. We are now, said Nigel, living with the unintended consequences of legislation. It is possible that industry did not engage soon enough, as legislators have a different point of view to those employed in this industry. ReacH is now three years old and was designed to unify the chemical management policy of the EU member states. Nigel kindly reminded us of the meaning of the ReacH acronym, and SVHC’s, which are substances of very high concern. SVHCs are carcinogens, mutagens, reproductive toxins; persistent, bio-accumulative and toxic. Legally, EU suppliers must respond to downstream users within five days for information on substances– in reality this does not happen. Only 50 per cent respond. The SVHC list is getting bigger – there 15 substances at first, there are now 38, with possible another 11 coming along, and new ones about every six months. Of course, you have to ask about information on the new substances, but this does not happen, either. The organisation called ChemSec maintains a list of 356 substitute it now (SIN) substances and the users of these which are now on SIN list have to find replacements, and when they do, they find that they are not on the authorisation list! ECHA say that 12,000 substances need to be registered,but only 4,400 have been so far, and many of the registrations were incorrect or incomplete.
Now there is imminent legislation that means ensuring that the materials you use do not come from an armed militia band deep in the jungles of the Congo, a further nightmare, and RoHS is already bad enough without recent developments that mean that WEEE and RoHS directives are being re-cast, or modified, and there is continuing uncertainty as delays are being made to the final legal text. That esteemed Belgian Herbert Von Rompuy has leant his name a further fudge known as “open scope” and more and more substances are being added – such as brominated flame-retardants, with little or no scientific evidence attached, as shown by the recent activities of no less a body than IPC. Oh yes, and CE marking is likely to become mandatory when new RoHS is being complied with. Just another cost for you in this competitive world (Figure 2).
October 7
Doug Pauls and David Hillman from Rockwell Collins duetted the way into the day (Figure 3) with a look at IPC specifications for cleaning and cleanliness. IPC-A-610 deals with visual cleanliness, J-STD-001E covers systematic cleanliness, IPC-5704 is for bare board cleanliness, and IPC-9202/9203 is for process qualification and materials compatibility. Beauty may be in the eye of the beholder, but it takes experience to know the difference, so it is best not to look at anything at a magnification of greater than 4×. With circuitry getting smaller and packing density increasing, cleanliness gets more and more important. Materials used during packaging, and the cleaning of a board with an eraser, are typical examples of the guidance covered by IPC-A-610. Contaminant is a warning, and every production facility should have a standard based on how much contaminant can be tolerated. Testing with ionic extract devices in accordance with IPC-TM-650 is recommended.
Also in 610, there are a number of useful drawings and pictures which illustrate the effect of flux residues, the damage that can be done with debris that has been left on a board and subsequently plated. White residues offer something of a minefield, surface appearance leads to the detection of material incompatibility. Understand materials, how they are processed, and what the results should be, urged Doug.
J-STD-001E – Section 8 Cleanliness – Doug chairs the cleanliness committee at IPC and has been involved or 25 years, so he knows the subject well, and was thus able to point out the problems which may be caused by customers being ignorant of what is required. What the manufacturer needs to do is have a cleanliness designator which determines when cleaning is required, or not. This is invaluable in boosting customer confidence. Materials and process qualification might need the old Appendix C test, but it is worth having. It can show up design faults, the wrong use of materials, and incorrect processing. This standard focuses on the entire assembly, but quite often residues can be localised, and cause failures, and the IPC Ionic Cleanliness Task Group is now pursuing localised extraction and evaluation methods. ROSE is only valid for process control, TM650 method 2.3.25.2 works better.
IPC 5704 – is a standard based on ion chromatography and the methodology is gentler than the traditional 75/25 process, and is very exact. The impact gives levels which are acceptable to the test laboratories as bare board cleanliness levels and is a distinct improvement over cleanliness testing by ROSE. OEMs have reported successes with 5704 and a reduction of failures attributed to cleanliness.
IPC 9202 and 9203 – intended as an international protocol and uses the IPC B-52 test assembly, a coupon for which was shown. IPC TH650 is the instrument of test, and IPC 9203 will cover how to setup a process to provide SIR under B.52.
For further guidance look at the IPC Cleaning Handbooks. For inspiration contact Doug or Dave.
About 17 years experience is also enjoyed by Ralph Hoeckle of Zestron Europe who was answering the question on why we clean if we have a no-clean process. His company provides cleaning solutions to the SMT industry, with water- and solvent-based cleaners. Based in the EU, the USA, China, and Malaysia, they can provide an analytical report on boards brought into their laboratories.
No clean means encapsulated residues, which can be guaranteed by a solid-soldering profile with a reasonable process window.
No clean does NOT mean that the PCBs used are “electrically”clean.
It does NOT mean that solder pastes will supply a “no clean”result under all circumstances.
It does NOT mean that other contaminants are not harmful.
No clean does work if a proper soldering profile is employed, but if there is climatic stress, and subsequent cracking, then corrosive acids can cause a malfunction.
Contamination accounts for 8 per cent of malfunctions of a circuit board;activators, rosin or resin, thixothropic agents, handling residues,fingerprints, dust, oil, and salts, can cause problems. After cleaning, J-STD001 indicates ionic contamination 1.5 μg/cm2 resin residues <4μg/cm2. SIR>8×10−8 ohms. To avoid electrochemical migration one needs robust circuit design, air ventilation and encapsulation or coating and adequate cleanliness.
Why clean? Remove activator and you avoid corrosion risk. Remove resins and you avoid conformal coating delamination. Cleaning removes fingerprints and dust. Lead-free solders need higher temperatures and have activators that need cleaning. Aqueous cleaning can involve spray, ultrasonic and spray under immersion. Semi-aqueous is ultrasonic and spray under immersion. He proved that cleaning can reduce the thickness of conformal coating by 25-50 μm. On 220 boards per day, i.e. £4,000 per annum. A handy saving.
The European Space Agency (ESA) sent along Dr Barrie Dunn, who was concerned about tin whisker growth. The next PERM meeting will be in Noordwijk in The Netherlands and will be hosted by the ESA. Back in 1972, they set up the first soldering standard, and the first tin whiskers were seen in 1974. In 1992, ESA prohibited the use of pure tin and things took a turn for the better. But other crystals were seen, such as silver sulphide, tungsten whiskers, aluminium whiskers from electromigration; and from 1998 to the present day, they have had to deal with a number of whiskers seen after thermal cycling, from −50 to+100°C. They have seen them in IC lead-frames, and in other critically important control systems. ESA have their own standards under the ECSS label. He illustrated how they had run tests using a C-Ring, coated with various metallic layers, 5-8 μm thick, fused in an oil bath, and had then measured whisker growth, of some 4-6 mm in length as an example. Fused tin-plating on all C-rings showed no whiskers after 15-and-half years. There is a great deal of activity worldwide on whiskers, with no less than six training schools in Europe. ESA have a comprehensive lead-free control plan which involves inspection, risk assessment, which is vital as they want a 30-year life for some space vehicles. Conformal coating can slow down whiskers but they can also grow them. In his conclusion, it was fused tin lead that gives an excellent solderable finish.
Jon Anderson from Humiseal spoke on creep corrosion, a problem that very definitely exists, and has done since 1955 which is when the first technical paper was published. Creep corrosion is the corrosion of elemental metal (Cu,AG, etc.) in the presence of sulphur compounds. It can be in a liquid or gas state, and generally requires levels of moisture. Oxidising can be assisted by halides. Silver creep corrosion does lead to field failures, and the use of silver in electronics is becoming more prevalent, as is silver plating. Sulphurous gases in the atmosphere exist everywhere but are more common in the emerging economies and in light industrial areas. Humiseal have run some experiments in which they took some boards which had been plated, coated,immersion silver finish. Using a panel with no coating as a “control”they tried various coatings on 72 test coupons, MFG at 25°C, ±75 per cent RH, 200 ppb H2S, ten days, SO2 100 ppb, NO2100 ppb. SIR testing was then run for 30 days at 40°C, with variable results. Extended heat and humidity did not affect corrosion, but UV and Silicone rubber gave better results, with UV showing almost no signs of corrosion, whereas the silicone rubber showed an unusual tarnish but no creep corrosion.
Summary: Use Humiseal Gen 1 Immersion Silver and use a UV coating.
Marion Quorrington of Cooper Industries works in harsh environments, but it does not show at all. Her company is one of the leading global manufacturers of control instrumentation, and over the last 20 years the move has been towards placing such instrumentation close to the processes that they serve. The human environment is pretty well known, but in a harsh environment it can be very hot and or very cold, or both, with high and low humidity; it might be near an EMF field, or in corrosive and explosive atmospheres; there may be severe vibration and shock. For them the material selection covers PCBs and components, solder and fluxes, the assembly process chemicals, and circuit protection strategy. They have to consider thermal properties, any chemical interaction between materials, any chemical interaction with environment, and circuit protection as a physical or chemical barrier. Smaller components may have less mass, but they may have higher TCE stresses, and interconnects are the largest single cause of failure under mechanical stress. Enclosures are needed to provide protection, in most cases. Verification testing can include HASS and HALT testing, EMC testing and humidity testing.
Typical environments can include marine, in which there are extremes of temperature, there can be salt-water immersion, controls must be free from RFI interference on marine radio bands, as well as withstand rhythmic vibration from ships engine, pounding and shock from waves and storms. There are well-established test regimes inc., Lloyds Register, DNV, and ABS. In industry,there will be chemicals, and/or corrosive gases, so protection is a must have,and here HALT and HASS testing is for thermal or mechanical stress. They have some control instruments in Canada, operating as wirelessly powered instrumentation on an Arctic pipeline, and there is little margin for error. Consider plastics at −50°C, let alone humans! It is going to get more demanding as electronics are needed in remote, inaccessible and hostile environments. At least they will now know who to talk to.
Andrew Clarke of AIM asked if it was better to have head on pillow or head in pillow. HoP or HiP? There are many solder joint issues – there are long lead times, coatings are getting thinner, many different alloys are used for coatings, the industry has by and large had to switch to lead-free, there are slower or longer wetting times, and the change to solder alloys used for BGAs. So, what to do? Develop new solder pastes, and fluxes for higher temperatures,that is what. So what is HiP? These are imperfect solder joints between component balls and the pad on the circuit boards caused by soldering or component failure, SR HIP is solderability related, and CRHIP is component related. The latter may be because the bump and ball devices may have been supplied out of true, or they can distort during the process. They can also lift, and oxides can form between the two surfaces (ball and paste), so even if the ball drops back the oxides spoil the connectivity. Solder paste deposit wets the pad but does not fully wet the ball, so there is no mechanical integrity. This defect usually appears when there is some thermal stress, or physical stress; namely, the person drops the phone onto a stone floor.
SAC305 has appallingly bad drop resistance, so BGA manufacturers have changed to SAC105 1 per cent Ag and 0.5 per cent Cu plus a small addition of magnesium,which give 1.5× greater shock and drop resistance. However, this needs 240°C in the reflow. Andrew explained how AIM had run a study on BGAs with known HiP problems; they were matt, spotted or shiny balls, all of which had significantly different surface structures under magnification. The addition of as little as 30 ppm of magnesium had a profound effect. The reflow profile did not have great impact on solder paste chemistry, but it was the BGA ability to warp which was the major cause. The balls which were matt performed worst and now AIM have some new solder paste chemistries with greater thermal stability over a wider range which has helped eliminate HiP problems.
The final speaker was Callum Drummond of Retronix who wondered if component refinishing/re-tinning was both good practice and reliable? This is all to do with the re-use of old inventory relating to their solderability condition,which is pretty much all of it. Old components should have a solderability test completed, as when they are disassembled, they can loose an arm or a leg, so to speak. That would be more than just a solderability issue, obviously. Causes of de-wetting of recycled components are contamination, oxidation and or plating issues. Cures include strip and re-tin, re-tin only, or re-ball. Processing would be done at Retronix in line with J-STD standards, and with BGAs they can refresh or replace the alloy. Reliable? From what Callum told us, it looks like it (Figure 3).
The SMART Group is for smart people. Putting the whole thing in context, the next time you go to a shopping mall that has some planes parked outside it, and you have fought your way through check-in, security, departure lounge, departure gate, air bridge, the aisle to your seat and stuffed your bag into the overfull overhead locker, you will doubtless be asked by the straight-faced driver to“sit back, relax and enjoy the flight”. Given the amount of professional care, knowledge and expertise that goes into the manufacture of what is called avionics, which will be in your “plane whatever the make or nationality, you can relax in the knowledge that you are not about to fall out of the sky because some component has dropped off. But you can if you want to;it’s head on pillow time.”
Well done, the SMART Group, yet another most successful conference, the industry is well served by your presence and the members are well served by events such as this.
J.H. LingAssociate Editor
