|
Your Equipment
A Brief Comment: We read in the Directive of the European Parliament and of the Council on the restriction of the use of certain hazardous substances in electrical and electronic equipment, given at Brussels, 8 November 2002 under 2000/0159 (COD) - C5-0487/2002 - PE-CONS 3662/02 - ENV 581 - CODEC 1273 in “Article 4 – Prevention - 1. Member States shall ensure that, from 1 July 2006, new electrical and electronic equipment put on the market does not contain lead, mercury, cadmium, hexavalent chromium, polybrominated biphenyls (PBB) or polybrominated diphenyl ethers (PBDE).“ where the term ‘put on the market’ needs some clarification as manufacturers of electronic equipment may not be clear about it and the implications it has for the target date of July 1, 2006. We enquired about it at the Ministry of the Environment in Berlin and received the ruling that ‘put on the market’ means delivery of the product to the end customer. This clarification makes it all too obvious that industry has to set their own target dates depending on the time it takes them to bring product from manufacturing to the end customer. Furthermore, it will be important to clear all intermediate product that contains lead out of the ‘delivery pipeline’ prior to this date or to ensure that such product is only delivered to customers outside of the EU where the directive does not apply and to customers who have not specified ‘lead-free’ product. By the way, manufacturers may continue to use lead in products that will not be put on the market in the EU even after July 1, 2003. Whether this exception will make things easier for industry has to be judged by the individual manufacturer.
Some Basic Considerations We already should have all the relevant information about the equipment: be it wave soldering, reflow soldering [infra-red, convection, condensation], selective soldering, tinning and coating equipment as well as hand soldering stations. We now have to ask the question whether they all are compatible with ‘lead-free’ soldering? But what does ‘lead-free capable’ really mean? We can be certain that there is an entire spectrum of opinions out there and your opinion might be tainted by the fact whether you sell or you buy such equipment. If you have not been very clear and adamant on many of the points when buying your latest piece of equipment you may have received a unit that the vendor considers ‘lead-free capable’, that however may not perform to your satisfaction within your lead-free process. You may have talked cross-purpose. Effects on equipment must be distinguished according to the type of lead-free solder to be used. Fundamentally there are two directions one can go: either the use of alloys with substantially lower melting point or alloys with substantially higher melting point. Those alloys with lower melting point (usually Bi-solders as Indium would be too expensive) also exhibit a much lower presence of tin (around 50%). Typically the tin content can be compared to that of eutectic SnPb-solders. All those alloys that melt at higher temperatures approach pure tin. Their tin content lies above 90% and in cases such as SnCu the copper at 0.6 % could be interpreted as an ‘impurity’. The high tin content makes itself felt especially wherever we encounter liquid solder. It is important to understand that it is the tin, which is the ‘aggressor’. It is the metal that causes the chemical reactions during soldering. The mechanism can be described as the creation of an intermetallic and then the dissolution. This leaching effect causes two problems in our equipment:
When examining our present equipment we have to keep in mind the alloy we want to use and the possible detrimental actions of tin at higher temperature and higher content percentage. However, ‘lead-free capable’ as a concept does not limit itself to material properties. It is very important to make sure that the equipment can meet all the other process requirements that are needed to solder with no-lead solders.
The Wave Soldering Machine A brief look at the flow soldering machine makes one realize that it consists in essence of these parts:
If we assume that we will use one of the low melting alloys with ‘normal’ tin content, there is little we will have to demand. Because of the decreased wetting ability of all of the replacement solders, we would like to see nitrogen capability in our equipment. The absence of oxygen improves the wetting performance. At the same time we reduce the amount of dross and hence our cost. A look at dross-data collected for the different alloys and the cost of these alloys is quite enlightening. Depending on the amount of bismuth in the alloy, the metal can actually expand when solidifying. There is a real danger of cracking the pot if no pre-cautions are exercised. These may entail either ensuring that the solder will never freeze (power outage!) or by using a solder pot that compensates for the expansion. Early experiments have indicated that the flux should be changed to something more suitable when using bismuth solders. Hence a look at the fluxer may be a good idea. Make sure that the fluxer allows for low maintenance and that we can control the amount of flux applied. Once you move to an alloy with higher tin content and higher melting point the entire situation becomes far more critical. We should have a detailed look at all the machine’s components:
Conveyor: Smooth operation, parallel rails and chains/fingers to hold boards securely are most important. Furthermore, the engineer designing such a system for lead-free solders must understand the implications of a) a higher melting point (thus even higher temperatures already in the pre-heating section) and b) a lower density of the solder. He must take care that there is no chance of rail bending or warpage when heated and that greases used on the conveyor are compatible with such higher temperatures. As the expansion of the PWBs will be increased by some measure (we assume that you follow the recommendation of the components manufacturer to limit the temperature jump from pre-heat into the liquid solder to < 100 K) the ‘springiness’ of the fingers or the solder pallet have to account for it. The lower density of the solder will affect the peel-back of the solder at the exit from the flow (wave). If approximately the same conditions as for lead-bearing solders are to be achieved, measures have to be introduced to compensate for the lower weight. This can either be a steeper inclination of the conveyor or some other means as for example a different nozzle design. A simple vector diagram quickly shows that a steeper conveyor angle improves the rolling effect of the solder peel-back in direction of travel. If we thus manipulate the gravitational pull we may calculate that angles between 7 – 9 º may be required. There are other properties that play part in the proper peel-back: viscosity, surface tension and naturally the flux. The big disadvantage of increasing the conveyor angle is the height difference between entrance and exit. The handling with periphery units, i.e. integration of the soldering system into a production line therefore might become a problem. To improve the flow properties of lead-free solder alloys without the negative influences caused by the change of the conveyor angle, SEHO was looking for alternative solutions. Instead of changing the conveyor angle, SEHO slopes the solder nozzle up to 2°. This causes a steeper angle resulting in a higher flow speed of the solder alloy which neutralizes the worsened wetting properties of lead-free solders. Compared to the angle adjustment of the conveyor this change of the solder nozzle angle has a decisive advantage: the wetting length and –time are kept on the same level without any time-consuming readjustment of the solder pot height. Additionally, new nozzle designs can help to keep the conveyor angle at the traditional slope. Nozzle geometries as for instance offered by SEHO, which show higher flow properties, are already used successfully in various production machines. As there is the possibility of ‘modified fluxes’ one would be well advised to ensure that the machine can keep these fluxes out of the conveyor system. Gumming up of chains and bearing is a high maintenance item.
Fluxer: An intensive search by chemists in their arsenal of chemical ‘weaponry’ have not discovered any more suitable activators than those we use at present. There are, naturally better chemicals, however, they are either far too expensive, too rare, hard to come by or toxic or have other undesirable properties. Hence we seem to be stuck with the di-carbon series of acids. With their point of disassociation of about 160 ºC, they are not really suited to meet the thermal requirements of the higher temperatures demanded by most of the lead-free soldering processes. But it is not only the melting point of the solder that makes their application problematic but the stringent requirements of many of the component manufacturers. During wave soldering it is not unlikely that the pre-heat temperature will have to be increased to a level that the fluxes have difficulty to cope with. Failure of fluxes would be pre-programmed if no counter measures are taken. The most common proposal for adjustment is an increase of the rosin content and a decrease of the solvent part. Switching the sequence of fluxing and pre-heating as we find it already in some selective soldering machines and some other applications in Japan, may have to be considered as another approach. One cannot help but wonder whether the move to lead-free will not also entail a certain re-awakening of cleaning operations. Such changes in flux make-up entail that the fluxer has to cope with higher viscosity and higher contamination. Proper exhaust and perhaps even a removal from the inside to the outside of the machine will ensure that maintenance is kept at a minimum. A separate fluxing module in front of the soldering machine offers various advantages. First of all, this ensures minimum contamination of the process area. Furthermore, removal of the fluxing module to the outside of the machine offers more room to enlarge for instance the pre-heat area which is a "must" in case the conveyor speed shall not be reduced.
Pre-Heaters: The function of the pre-heaters has two major aspects: measured temperature increase (gradient) of the assembly and achieving of a targeted temperature prior to entering the liquid solder. Nowadays, thermally complex assemblies are heated with a gradient of between 0.5 and 2.0 K/sec. Making a few basic assumptions we will run through a very simple calculation:
Under these assumptions the temperature increase amounts to 120 K. If we use a gradient of 2.0 K/sec we have to dwell in the pre heat for 70 seconds and for a gradient of 0.5 K/sec this augments to 280 seconds – not including a ‘plateau’ should that become necessary to evaporate solvents (e. g. water). The chosen conveyor speed of 90 cm/min translates then into a length of required pre-heat of 105 cm, however for a gradient of 0.5 K/sec the total length of pre-heat needed would be 4.20 m. Not many pieces of equipment would meet that requirement. As an alternative we could reduce the speed of the conveyor at least in the pre-heating area (split conveyor) so that we do not negatively affect the contact zone in the solder. Reducing the conveyor speed may be possible in many applications, however if no slack time is available during production it would automatically translate into a reduced through-put. This is of course in most cases not acceptable. First of all you should check whether the configuration of the pre-heat area can be changed, e.g. by installing more powerful preheat modules (quartz or convection instead of IR). By taking out the fluxing module into a separate module, the pre-heat zone can be enlarged.
Solder module: The most critical part in a wave soldering machine is the unit where the liquid solder is contained and pumped. This is particularly true when we consider lead-free soldering with SnAg, SnAgCu or SnCu alloys. The normal 305 or 316 steels are not resistant enough to cope with the aggressiveness of the tin at elevated temperatures. This has been known for some time as companies had experienced substantial problems in hi-temp applications. Particularly those parts that are subject to movement and possibly friction suffer the most. Pumps, impellers and some parts of the wave formers show very early deterioration. But certain areas of the walls also get attacked. There are some metals that seem to be impervious to leaching by tin (e. g. Hastelloy) but such material – it is proprietary – is expensive and extremely difficult to weld and cut. Thus the manufacturers of wave soldering equipment concentrated on coatings for their pots. What we see ‘out there’ in the different pieces of equipment shows the entire range of experiments – some more successful than others. It becomes obvious that many of them have only been tested under laboratory environments as they may stand up to the task if they are left undisturbed. However, in real life, the pots and pumps are abused a lot. Scratching with screwdrivers during maintenance, knocking pumps against concrete floors to dislodge stubborn dross particles are common practice. If we account also for such treatment, there are few coatings that would endure the two-pronged attack of inconsiderate treatment and leaching by hi-temp tin. Even the best coating will be ‘under-eaten’ if it is cracked or chipped. The pot-treatments range from simple painting with Teflon (not to be recommended) via impregnating with titanium-nitrate to ceramic coatings. An in-depth discussion with the manufacturer is highly recommended. SEHO developed together with industry partners and research institutes a special composit coating. This protective coating meanwhile is successfully used for several years and is suited for all known lead-free solder alloys. The increased cost of dross will justify the use of an inert atmosphere by itself. Working without dross reduction measures would be too expensive. Hence the need for an efficient inerting system. The improved wetting performance under nitrogen will then come as a fringe benefit, although absolutely recommended even for other lead-free soldering processes. Without nitrogen one would have to go to even higher process temperatures, something the materials used would not forgive easily. A final word about pot size: the larger the pot the easier it is to generate a flow with little turbulence and greater depth. It also will take longer to exceed dangerous contamination levels. However, filling a large pot will be much more expensive as lead-free solders can cost ‘a pretty penny’. The company has to decide where their optimum pot size is to find.
Cooling: Check the possibilities to cool the assemblies directly after the soldering process, i.e. before the assembly leaves the soldering machine. An effective and fast cooling results in a very positive influence on the metallurgical structure of the solder joint. Metallographic sections through solder joints which were processed without any controlled cooling, show for instance a distinct intermetallic phase (Cu6Sn5). The structure of the solder does not show a homogeneous solidification.
Reflow Soldering Systems We distinguish three of the best-known processes: Infra-Red Radiation, Convection, and Condensation heat transfer. Infrared Systems There are few movable parts in such equipment. However, attention must be paid to the conveyor and maybe improved insulation. The basic problem why the infrared technology was moved off the market was its inability to provide small ΔTs on the assembly. Potentially this inadequacy could become more pronounced, as the temperature will have to be increased. Furthermore, many such systems are relatively short and have problems already in SnPb processes to maintain specific profiles over longer periods. It is unlikely that the situation would improve for alternative alloys. We have to question again whether the ΔT will play the same important role in lead-free soldering as it has in traditional solder joining. If not, the argument reduces itself to the large temperature difference between the heater source and the heater target. That infrared equipment also will need the feature of inert atmosphere coverage is obvious. It is never the soldering method proper that reduces wetting but rather material contribution. Since the new pastes will be based on somewhat higher rosin content, managing the tunnel’s atmosphere to minimize soiling of the interior, is highly desirable. Convection Systems Although ‘reflow’ is only a ‘managed heat transfer’ there are still many pieces of equipment out there that do not meet the requirements of lead-free processes. When thumbing through the maintenance record of one of our customers, who is still running SnPb processes, I noticed that the ventilation unit in the peak zone of his systems had to be replaced every three months. This had become routine as a preventive maintenance procedure. The motors and bearings were not able to cope with the peak temperature for longer periods. If machines have problems during SnPb processing, we can only fear that at 20 – 40 K higher peaks the system is close to constant shut down. Again it is the moving parts (motors, bearings), which pose the problem. And even if the manufacturer of the system provides replacement parts free of charge during an extended warranty period, the situation is not solved. Because it is not the cost of the repair but rather the down-time of the line, that costs money. International companies calculate the down-time of a major mass production line to be around € 10,000.00 to 15,000.00. As one of my engineering contacts in such a company once put it: “after 10 hours of downtime the company can not only buy a better machine but I also can find myself another job.” Special attention therefore should be paid to high-quality blower units and motors which ideally are side-mounted, since top-mounted parts would be located in the hottest machine area. Simultaneously, increased process temperatures require a very effective energy transfer. Heating zones which are not only located in the upper machine part but also in the bottom area, combined with an optimized air or gas circulation, as for instance in case of tangential fans, ensure an effective and component-sensitive heating of the assemblies. Therefore, setting of low oven temperatures is possible which especially in case of lead-free soldering processes is of importance since the temperature-stress on the components will be reduced to a minimum. At the same time the oxidation rate will be kept as low as possible. Systems with so-called multi-peaks, i.e. double- or even triple peaks, moreover ensure a very flexible temperature profiling. Nitrogen capability is another requirement. The worsened wetting situation must be compensated by nitrogen and not by even higher temperatures. Low consumption and a reasonable level of residual oxygen have to be targeted. Consumption can be lowered and resources and money can be saved if reasonable ROL values (100 ppm in the peak zone) satisfy the process requirements. As pointed out repeatedly, we expect higher rosin content in pastes and hence insist on better flux traps and gas management methods within the equipment. Stalagmites and stalactites look pretty in caves but not in convection tunnels. Not only do they ask for serious maintenance but they also pose a danger to the cleanliness of the assembly. The soldering system therefore should be equipped with a multi-stage condensate management (infeed, heating zone, cooling zone and exit). A filterless flux management system provides some advantages since filters may gradually clog and thus influence the process negatively. The new temperatures will be close – if not above – the glass transition region of the laminate material. We expect more problems with warpage and deformation of the PWB – and that is what is seen already. Once the PWB has lost some of its stiffness, it has to be supported. An adjustable center support can often help greatly, if the layouter has accounted for such features. At the exit cooling defines the crystalline structure of the joints. We still lack information on desirable cooling rates. But we do know that the replacement alloys can create different crystalline structures depending on the cooling rate employed. Research presently tries to identify the ‘best’ crystalline structure and the cooling rate that can achieve it. As we expect the results to be available soon, we require adjustable cooling possibilities at the exit of the machine. Condensation Systems In condensation reflow systems the reflow temperature is most often (at least in saturated vapor systems) defined by the boiling point of the liquid in use. For alloys with higher melting points the liquid has to be changed to match boiling point with melting point. Research carried out at the Technische Universität Dresden (Prof. Wolters – personal communication) seems to indicate that a lower ‘super heat’ (distance: melting point of alloy to boiling point of liquid = process temperature) may be used than in other reflow processes. The reason may be seen in the very low level of residual oxygen present in the vapor. As there are many different liquids available with a large range of boiling points we do not see any problems arising from special requirements posed on chemicals. The thermal profiling of condensation reflow is largely controlled by the conveyor speed. Changing the heating capacity is only used for some pre-heating technology. The question whether such systems are lead-free compatible thus seems to limit itself to an assessment of the heating capability and whether the increased energy requirements of the higher boiling point liquids can be met. Although the liquid has only a limited ability to dissolve rosin, it would accumulate in the sump if it were nit filtered out regularly. For the new pastes better or more frequent filtering may become an issue. This is a question that is maintenance related only as the vapor, being a distilled phase, is clean even if the sump has been contaminated. Today’s equipment is well insulated and though the temperature inside will be increased by 10 – 20 K, the skin should remain within acceptable limits. In most machines the conveyor is ‘basket like’ and does not require any additional support structures, except perhaps for double-sided reflow. What has been said for the crystalline structure when discussing convection reflow equipment also holds for condensation reflow systems. Cooling rates at the exit have to be ‘settable’, otherwise we may be left without the means to produce ‘optimal’ solder joints.
Other Soldering Processes From the discussion of the main soldering processes one may glean those aspects that also apply to other operations. The derivation of necessary aspects is straight forward. During selective soldering either liquid solder (mini-waves) or reflow methods (hot gas or Laser) are applied. However, all aspects described before, are less problematic as all processes are only selectively and do not affect the complete assembly. Especially in case of miniwave soldering, lead-free processes do not differ much compared to the traditional SnPb soldering processes, provided ideal peel-off angles can be achieved with flexible handling (gripper) systems (e.g. tilting, turning of assemblies etc.). In those cases, the changing flow pattern of lead-free solder alloys are not relevant. From a materials point of view suitable protective measures (e.g. for solder pots, pumps etc.) have to be considered with regard to the aggressiveness of lead-free solder alloys. The situation is different for hand soldering. The life of the tips of soldering irons will be reduced substantially. And, if several alloys will be in use, the control of the proper tip temperature [melting point of the alloy + 70 K for the heat-sinking experienced during soldering + approx. 35 – 50 K ‘super heat’ to ensure proper flow and wetting] will be more difficult than in recent times. The difference in melting point would ideally translate into different tip temperatures for each alloy. Maybe the solder station vendors could consider making life for the supervisors and operators a little easier by color-coding the different pre-selected settings?
Concluding Remarks During your assessment of the equipment you certainly came up with a wish list that you can compare to your present situation. Now you ask the questions:
|