Emerald Group Publishing Limited
Copyright © 2008, Emerald Group Publishing Limited
EIPC Conference Rome 24 and 25 January 2008
Article Type: Exhibition and conferences From: Soldering & Surface Mount Technology, Volume 20, Issue 2.
Day 1 Thursday 24 January
The many delegates were welcomed by Rex Rozario, the EIPC President, who outlined some of the events that are to take place in Europe and elsewhere during the coming year. He reported that in China Somacis-Graphic are now producing 30-layer multilayer boards with 97.8 per cent yields after 4 months in production, and flex/flex-rigid boards are also in production. They will break- even in the next 3-4 months, so within the first year of manufacture. Well, that is how it used to be in Europe. But, Rex was confident for the future in spite of a shaky start to the year.
Russ Crockett from Du Pont was the moderator for the Management and Technology overviews. Anthony Walker from RTC North talked about ProSurf under Framework 6 Coordination Action, promoting innovation in the surface finishing and printed circuit board manufacture industries. The web site is available to the industry, and may be found on www.prosurf-online.eu. Anthony talked about partnering opportunities under Framework 7. New knowledge, improving competitiveness, and developing the European research area were amongst the aims.. The structure was clear people, ideas, cooperation and capacities. A total of e50 billion is available over next 10 years, and interested parties will find a¼n organisation called Cordis is the main route into FP7. ETI support projects and a National Contact Point are in every country; there are also regional contact points. But, the main point is to help yourself, work with people you know, in an industry you understand.
Bernard Bismuth spoke on behalf of FIEN, who were founded in 2002. Not a trade union or trade association, they are more of a think tank. Membership is 1,100 companies, and 50 per cent of their production is for export. It is less to do with protection; it is more to do with showing how electronics can play a major role in the future needs of society. The EC does not seem to include manufacturing in it's thinking even now, so Bernard would like a cooperation that can bring ideas in this field to the commissioners. The markets that will be served will be smaller (cars and telecoms) but bigger in the industrial medical and aerospace sectors. Other new opportunities lie in power savings for the home, in transport, in industry, and individual well-being, at the home, and in medical care. PCBs represent 12 per cent of the components sector, and are therefore vital. Bernard made it clear that the USA, Japan, Taiwan and Korea are each three times bigger than Europe, but they have only one language, against 25 in Europe, and in Europe we all compete with one another. What is needed is a way in which we can develop products which can be manufactured, not elsewhere, but in Europe.
Bernard is passionate about this, the need for Europe to be careful. If we lose all manufacture out of Europe, we will lose all R&D as well; but how we supported manufacturing was not entirely clear. Innovation in the process is as important as innovation in manufacturing. And on the subject of innovation, why were there so few manufacturers at this conference? Considerations include language, time, cost, staff reductions. But, suggested Bernard, why not have manufacturers supported by the EC to attend seminars such as the EIPC ones? We should talk to the people in Brussels. Mais oui. One for all, and all for one, was his mantra. Flawless rationale, a difficult route internationale.
Dominique Pellizzari is the boss of the CIRE Group, and wondered what can the European PCB industry do to get stronger? Some facts and figures first the French PCB industry represents 9 per cent of European production, it is worth e234 million, and comprises just 31 companies (down from 68), and they are the 5th largest producer in Europe. Most members are small medium companies, serving an e800 billion electronics industry in France; challenges include a reduced supplier base, restrictions on number of hours, tougher environmental regulations, and changes in the “value chain”.
The structure of FIEN was described, with Gixel, FIEEC, and a competitive cluster, such as in Toulouse, where the companies and universities specialise in aerospace with technology companies gathering together in serving one industry sector. Insurance, payment terms, industry standards and regulations were all matters which came under FIEEC and FIEN takes on the promotion in all aspects, lobbying to promote R&D. Translate such a national approach into a European one was the point made by Dominique. In an ideal world, Dominique.
Giacomo Angeloni from Somacis in Italy had a swift look over his shoulder. Well, there were 767 PCB manufacturers in Europe in 1990, now there are just 342. The PCB market in 2000 was e4,775 million, and this has dropped down to 2,748 million in 2006. In Italy, the number of PCB companies has dropped down from 167 to 33, and there are 25 per cent less people employed in the PCB industry in Italy than there were before.
What manufacture we have left is paying higher prices for most of its supplies. We are paying about twice as much for petrol in Europe as they do in China. In Italy, motorists are paying $1.56 per litre, against $0.69 per litre in China. Looking at what the EC does for industry, the answer appears to be not a lot; 45 per cent of expenditure is in agriculture. But surely, we should be investing in technology, not crops. Technical economical and political actions are needed; we need European funds for research, for it is on the back of innovation that we shall survive. Much team work is needed here, said Giancomo. Like Bernard, Giancomo is looking at the EC for assistance, but it maybe that the commissioners have to be persuaded that the future lies in technological innovation, not olive trees.
Dr Udo Bechtloff from KSG Leiterplatten GmbH said that his company has the same problems as anyone else, but they have been successful in their own right. It is a family-owned company, with a turnover of e9 million in 1996, which had risen by 2007 to 47 million. Founded in 1956, KSG lies in the former East Germany and before reunification were employing 3,000 people, but they aim to be down to 430 by 2010. They have made a e50 million investment in plant and buildings, serving 535 global customers, but through German “mother” companies. Growth had been a process of steady increase, and their product spread is 39 per cent for industrial electronics, 31 per cent for EMS, 14 per cent automotive, 10 per cent consumer electronics and telecom 5 per cent. With 25,000 orders per annum, they handle 100-200 different jobs per day, with an average order size of 1-2m2, with 96 per cent delivery reliability. They have grown in a declining market, admirably, where a market in Europe has reduced by 47 per cent in 7 years they have risen by 88 per cent. Supply what the customer wants, said Udo. Yes, but it just maybe that KSG started with a lower operating cost, and benefited from a lot of government assistance. Their technology road map was very interesting, and as was the fact that they fund ten R&D engineers out of turnover. Reliability of products, says Udo, is the most important thing, and reliability testing is carried out as a matter of routine. We suspect that a little modesty was being displayed here, with inspired management and strong staff motivation being key factors as well.
Neil Chamberlain of Polar Instruments discussed the design considerations for the selection of high- speed laminates. The choice of laminate for PCB production is now becoming critical to the success of the finished design not only for its ability to withstand lead free processing temperatures during assembly, but also to ensure that the electrical requirements of the design are met. The latest high-speed digital design techniques are now demanding superior performance from the laminate than standard FR4 can deliver. There have been many low loss and high-speed materials developed for the market but all at extra cost to the design, which must be managed carefully, said Neil. The decisions now need to be taken at the design phase of the project to ensure reliable electrical performance, which means that ownership of the laminate selection must be held by the design authority and managed carefully throughout supply chain.
Use of an accurate field solver that has the ability to alter transmission line losses with dielectric constant changes in the material is essential, in order to understand the limits and capabilities of the material selected and use of a standard electronic format for layer stack-up communication, designed to cross the bridge between design and fabrication requirements.
For the technical amongst us, Neil explained that there is an easier route for designers and fabricators to take, this ensuring clear consistent communication of error-free layer stack-up information. Best modelling requires good data in, and in conjunction with a good field solver, most issues can bet met. He commends the IPC Technet e-mail forum, and that of the IPC Designer Council. Or have a chat with Neil firstname.lastname@example.org
In a highly detailed technical paper, Erik Biehl from Multek in Germany talked about their lead-free programme where base materials were assessed for compatability and reliability. Thermal robustness is the key. They have a global programme to test lead free compatability and this is being done in both Europe and China. In the lead-free assembly process and with repeated reflows @ 2608C£3, the key parameters are the Td, the Tg, Anti CAF (electrical integrity) and low- modules resin. They have found some preferred materials by a process of elimination testing, using panels that have dedicated coupons for testing CAF, reflow and solder shock, TG and Td testing, and for through hole, buried holes, and standard materials. Highlights of testing to T288 included stable Tg (no degradation) on all Novolac systems, but delamination on T288 Td testing was noted, along the resin-copper interface. In his summary, Erik said that no delamination was found below 7x reflow on certain materials (Materials A-H. and A-E looked promising, whatever they were) and their recommendations were given in a detailed table.
Dr Wolfgang John comes from LPKF and shared with us his method for producing 3D PCB using laser technology. Here, the production of sensor packages is pertinent, and their application is myriad. Miniaturisation, increase in functional density, and a reduction of manufacturing costs all play a part here. Moulded interconnect devices can be produced in two ways:
1 two-shot moulding, which is limited, slow, difficult and expensive; or
2 laser-based technologies, which offer a long-process chain, and incur the need for chemical etching, and expensive processing costs.
So what now? The solution is a full additive laser process that combines the advantages of the laser with the advantages of a short-process chain. Thermoplastic materials doped with a metallorganic complex, or a laser- activatable additive is the basic building block, and the laser generates a micro-rough polymer surface for the anchoring of additive copper plating. Microline 3D Industrial laser system was shown, where 120mm lines with 150mm spaces can be made. The process sequence was described, just 12 simple steps, from ultrasonic cleaning, through to electroless copper, electroless nickel, and electroless gold applications. Reliability tests had given excellent results, especially for automotive requirements.
L.C. Chen from Du Pont in Taiwan proffered his views about the HDI process, challenges and recommendations. 50mm lines and spaces are now being asked for, so what are the challenges? He looked at the process flow; 50/75 is mainstream, but the reality is in fact for “cell phones, where major production is in the 75-100mm range but moving to 50/75 within the next 3 years”. For standard PBGA, the gap is even smaller, with 30/50. The challenge is etchback. The etching factor is worse with fine spaces; the copper thickness needs to be reduced to get a better etching factor, but the trade off is when you reduce it at inner layer non-contact exposure here is applicable LDI. It eliminates artwork, reduces cycle time, has production flexibility, and saves film, artwork and chemistries. LDI has 300 lines installed worldwide, with a prediction of 3,000 lines installed world wide within 10 years. The Off Contract Exposure Stepper is already in operation in Asia, and 20-40 units are under evaluation trial run production, mainly used for solder mask, supplied by Asian equipment manufactures. The process summary was given for classic, cellular and PBGAS and FCBGA production, which showed significant variations. Fine line etching limitations were explored by Mr Chen, who concluded that thinner polyester coversheets, with improved lamination equipment, is the key for HDI processing.
Dominique Pellizzari of CIRE moderated the afternoon session, which was always going to be about drilling. Stefan Kunz is the MD of Schmoll Maschinen GmbH, and presented a paper on drilling and routing for European PCB manufacturers. Europe is different Europe means high-end products being produced in a very short time, so cycle times are important. At Schmoll, they develop their own spindles, from 3,000,000 down to 10,000rpm, for many functions. They use 1/8in. shanks, whereas in Asia they use 2in. shanks. European machines should be very stable whilst running, with online tool breakage detection. Also, tool wear must be measured, so on their machines during each drill step, the head is measured. Stefan went on to describe the wonders of the Modul system, that reduces the footprint needed whilst increasing efficiency and offering an automated operation at the rear for loading and unloading, but full manual operation where needed at the front. Thus, each machine runs independently of lot size, with high-machine utilisation, and minimal manning levels.
Uwe Lenz is the MD of Ernst Lenz Maschinenbau who knows about the new developments in drilling and routing machines, as he builds them. Shorter runs lengths, faster turn-round times, pressure on yields, all contributed to demands on machine manufacturers, along with demands for greater accuracy and productivity. To meet these demands, they have a dedicated machine with spindle switch over, and controlled depth drilling. At Productronica last year, they showed off their new machine for drilling and routing aluminium with a cooling system in the drilling head which, well, cools things down. Multi-head micro- drilling machines have now become a reality, with a new Lenz machine in this sector, available with a pressure foot system which better drilling accuracy along with better hole wall quality. 30mm holes are now being requested as standard, he thought, here their DLG 560-6 machine will fit the bill.
Dominique wondered where the line was between laser v mechanical drilling. Uwe explained that a laser drill is just for drilling blind holes. For real holes, you need a mechanical drilling machine.
Jürgen Skrypczinski from HAM GmbH wondered about the quality and productivity when drilling and routing filled laminates. The heat concerns needed an undercut drill, with a helix-angle of 348. At HAM, they have tested entry material with coded aluminium from TMT, and have found that this cleans the drill much better and reduces a build up of aluminium on the drill. Some testing was done with various laminates, such as phenolic hardened, which gives less tool life compared to halogen free and standard FR4. In a highly detailed paper, Jürgen imparted a great deal of invaluable technical information based upon tests on various laminates, and HAM have a list of the best results and recommendations that will meet all drilling requirements.
X-Ray tooling was the subject of a presentation from Paul Waldner, MD at MIE. A good image recognition system seems the best way to start with an X-ray system, and Multiline have their XRT System which incorporates just that. Here, 1, 2 or 4 X-ray cameras examine the target pad, and the vision system calculates the centres automatically. A number of holes are then drilled. Any bias required can be added, and holes can be moved around as necessary. MIE have their own software for layer analysis, which can analyse registration on a layer-by-layer basis, analysing shift, and in addition to the stack up targets other targets can be added, and this software allows you to look at the board from different ways. Their Layer Analysis Package (LAPe) allows the customer to collect data for SPC purposes, so it is in some way a complete factory solution.
Pietro Zulli from Pluritec and Andrew Kelley from Xact PCB came to Rome to talk about how they have worked together to improve multilayer registration yield using X-ray drilling systems. The difference is that they have a system that measures the pattern deformation caused by distortion and drill a compensating series of holes that re- aligns the layers. The panels are given many targets along the edge of the panel and the data, once collected, can be used to determine the drilling pattern. The machine used is called an “Inpecta HPL”. Andrew went on to add that the data visualisation and analysis for the process engineer can assist in the identification of random misalignment, and he can predict linear or nonlinear scale factors and optimise accordingly. The Gemini-X CAM Interface can accurately predict scale factors and remove the need for internal prototyping.
Markus Di Marcoberardino is with ESI GmbH in Munich Germany. His topic was the UV laser process and quality requirements for glass reinforced PCB materials. The U/V laser process has many advantages, but a smaller process window along with a need for a sophisticated and well defined parameter qualification process join a slow drilling process BUT there are now hybrid laser systems which have a CO2 and a UV laser source on board; in the first step the UV laser opens the top copper, and the second step removes the dielectric. This is however unusable for through hole drilling. So, for drilling of glass reinforced materials, Markus steered us down the route taken by ESI; undercut is a problem, as is fibre protrusion and etch back, but otherwise all appeared to be well.
Thomas Michels owns TMT. Founded in 2003, TMT is a German company specialising in drilling materials, amongst a range of others. Entry and back-up material are his speciality, and in this field he started with the evident truth that hole sizes are getting smaller down to 0.075mm. Buying the cheapest is not always a sound idea, given the tasks that entry material has to undertake. He described the range of entry materials that are available and why they are used and where they are used. ALE is a lubricated one, APAE is an aluminium sandwich with paper in the middle; and LAE is buffer material over aluminium foil, without lubrication. Clean materials give better hole wall roughness. Cleanliness is next to godliness here with entry material as dirt can cause deviation of hole direction and distortion of hole diameters. A planar smooth backup board with a very hard surface reduces burring, and back-up boards of pressed wood show good results when comparing price and effectiveness. For micro drills and multilayers, a coated backup board is recommended.
Day 2 Friday 25 January
The first session of the day was opened with a paper from Michael Herkommer, of Umicore/Taiyo in Germany. Michael led us through the product range of Taiyo and explained how the decreases in BGA ball pitch, higher ball counts and thinner substrates have necessitated the development of new liquid and dry film coatings for the new generation of packages. They have a new range of LPISMs for wire-bonded packages, with improved migration resistance, and crack resistance, and a new range for flip-chip packages, halogen-free, with extra-low solvent content. Resolution is stable at 20mm, with flatness and plugging ability evidenced. In the dry film range they have one for ultra-thin chip scale packages, and one for LDI patterning, based on a new tough resin with lower CTE and high Tg characteristics.
Dr Manfred Suppa is the well known face of Lackwerke Peters, and being head of their R&D means that he is privy to the developments at Peters with photoimageable solder masks for electronic assemblies. Electronic assemblies have a lot to cope with temperatures, thermal shock, vibration, moisture, e-corrosion, salt spray, noxious gases and migration. Operating temperatures in motor cars where electronics have to perform are anything between 200 and 1208C, and these electronic assemblies are now using lead-free solders with multiple soldering up to 2608C. So, failure of an assembly can be down to chemical degradation, thermal degradation, and thermo- mechanical degradation. Dr Suppa led us through a close investigation of the performance of solder masks after temperature loads, thermal cycling, and with the help of a Dynamic Mechanical Layer Analyser, embrittlement and mass loss. Using the Arrhenius equation, Peters set out to look at failure due to thermal degradation, and with the AIF project, they are investigating the aging process under oxygen and nitrogen to maximise TE values. Happily, Peters resists live up to their name, and are resistant to just about anything that is thrown at them.
Anders Ekman came down to Italy from Sweden, where he is the Marketing and Business Development Manager for Huntsman Advanced Materials. Anders looked at the many ways in which the use of his company's products would be of assistance to the PCB manufacturer who faces up to the demands of registration accuracy as well as imaging properties and solder mask clearances. BGA fine pitch components are 0.4mm now, soon they will be 0.3mm pitches, and in volume production. So, the use of LDI would seem to be the best way to cope, offering high-digital accuracy whilst avoiding artworks, always expensive for prototypes. Huntsman have developed a LPISM for LDI work, available in semi- matt or gloss finish, which processes four time faster than their older products, with better energy and wave-length cross-linking at 355nm, and high resolution on small features, straight sidewalls after development, and mere 25m on the laminate itself.
There is always the risk of appealing to the chemist in us all, but not all of us are chemists. However, the highly detailed papers from these three speakers reflected the depth of knowledge, experience and expertise of companies who have served the industry well for many years.
The final session of the conference was moderated by Giacomo Angeloni, of Somacis pcb industries. Why not, we were in Italy. Alexander Kasper from AT&S AG in Austria joined with Dr Steve Jones to talk about industrial inkjet processes for PCB manufacture. Legend ink is already up and running, solder mask is on the way, but etch resist for inkjet is not yet available for production scale PCBs, and conductives, the ultimate goal. AT&S are looking for the ultimate low-cost PCB, and here they are pursuing new technologies, they do not to buy in a fixed system, but to devise one that suits them. Ink Print Head Machine and the Process are the key elements. The application has been identified, the business plan and cost calculation has been done, now they are tackling the technical problems, partnering with Printed Electronics Ltd They meet every 1-2 weeks, and have nicely short-communication paths. A good ink formulation is almost in place; they have to integrate a machine, the process will be developed by PEL, and AT&S will be the b site for it. And all the better for it, too. We await further news with great interest.
Dr Steve Jones spoke on behalf of Printed Electronic Industries, of Cambridge. PEL now have an industrial process it has been designed to work in a rugged environment, can print at industrial speeds, this is an iTi machine, which can take six different heads, Xaar heads mainly, for high-speed printing with a small number of heads. Using high- speed photography, they can see how the ink is leaving the head and how it is forming on the substrate, PEL know the PCB business and thus are a suitable partner for AT&S. They have an etch resist, and the initial work looks most encouraging. The future includes nano- conductive inks, multilayer dielectronic structures, and components. A panel per second is the goal.
Uwe Altmann is the man from Orbotech who we all know well, and New System was the company that we all knew well who were bought by Orbotech. So, they are now firmly into the inkjet field, and drop on demand is the future, a process which Uwe described. Drops come in a continuous stream, the viscosity is controlled by temperature (solvent evaporation), and the system boasts lower consumable and maintenance costs than before. Inkjet resolution is 1,440dpi. UV inks are attractive, no solvents, and with them you can create 3D structures. They also cure easily and quickly. The advantages of inkjet versus screen were clearly shown, and probably already well known. Conventional solder mask process steps against ink jet application were also shown, and here the speed of UV assists greatly in the drying process for legend inks.
Legend inks exist now, to come will be etch resists, chemical milling, solder masks and solder dams using inks only. Conformal masks and conductives are also in the future. They have done some work on 100mm lines and spaces using inkjet. They need 1,440dpi for this.
Gregory Blake kindly crossed the pond from the USA where he set up the Printar operation, to talk about advanced inkjet technology. Cost pressures are enormous on PCB production, so inkjet technology offers potentially attractive cost advantages for certain kinds of circuitry, but not all. Printar have 50 systems installed worldwide, and have 70 staff employed, dedicated to inkjet and nothing else. Having their ink approved is a key element here. Legend printing was the start, and here they have been successful; the advantages of inkjet v screen printing were underlined once more, including the availability of “serialisation” of online numbering. Gregory highlighted some of the problems with conventional solder masks, and the potential advantages of ink jet solder mask; here inkjet replaces lots of machines, people, and saves a lot of time a 2min cycle time against 133min! One the face of it, it sounds good, but one must ask the question, given the need for approvals and the meeting of specifications will it work with all solder masks?
Ashok Sridhar from the University of Twente in The Netherlands is working on his PhD researching the interface between inkjet printed semi- conducting components and the substrates used in the PCB industry. His technical paper covered inkjet printing using functional inks such as PEDOT and a silver nano-particle ink, on pcb substrates and glass. They have characterised adhesion between printed layer and substrate. Using a Microfab JetLab4 inkjet printer, Mr Sridhar clearly showed how the droplet size can be strongly affected by temperature and humidity. They have also used a silver ink on FR4. The problems and challenges are not a few. One is the absorption of the ink into the substrate, so they have measured this with an autoperosity measurement system, and they can see how problems might occur where two lines meet inside the substrate. Future work looks extensive, and the results so far look most encouraging.
Rob Haslett is the CEO of Patterning Technologies Limited. He showed the shopping list that is required: a system that produces high quality at low cost, is fast, handles complex materials, in small batch sizes and with full traceability. ereehere Here, inkjet technology can be of real assistance. If offers minimal set up time, minimal operator skills, fewer processes, lower cost, and with variable data. Getting rid of the wet processing is desirable, but how do you go down that route? With inkjet there are many parameters ink, print heads, print platforms, and print strategy they all have to be balanced. Factors to be considered are adhesion- surface wetting; viscosity and jetability, coalescence; resolution, reliability, speed and accuracy.
There is 10 years experience at PTL, a technical consultancy that is experienced in the bio-medical and flat panel displays, and etching applications as well as PCB. They have partnered with KLG in Germany, who know all about X Y positioning, and together they have launched JetRitew as a system. PTL are looking at the marketing, KLG are producing the printing platform, and the manufacturing and sales and support, which is now in place. Unsurprisingly, the printing platform is very similar to a drilling machine. PTL are providing the printing head, the ink, managing the suppliers, and notifying the end-users.
So far, they have a machine that has one printing speed at very high quality a 17s print time. With a 14 picolitre drop size, they are obtaining 100mm lines and spaces. There are two self-aligning print heads, one for white and one for yellow (legend inks). It is a compatible partnership and doubtless destined for great things interesting to see how his company has arrived at the same place, but via a rather different route.
The EIPC, as a vehicle for the dissemination of sound industry information, remains one of the “classic cars” of the PCB industry. Familiar, reliable, fascinating, and it is always a pleasure to be a passenger. If one is to have a conference in January in Europe, there would seem to be no better place than Rome. If one wishes to attend a PCB conference, there is no better place than an EIPC event. The next EIPC Conference is their (early) summer one, and they are holding this in Dresden on 29 and 30 May.
John LingAssociate Editor