SMTAI 2007 SOLDER JOINT AND SOLDER FILLET OPTIMIZATION FOR SURFACE MOUNTED TERMINAL BLOCKS Michel Hodak Dubravka Kusmic WECO Electrical Connectors Inc. Montréal, Québec, Canada [email protected] [email protected]

ABSTRACT The appearance and strength of surface mount solder joints are the objects of concern and scrutiny in all OEM and CEM involved in these processes. Management's profitability and productivity concerns often impose excessive and almost impossible demands upon line speeds and set up times. These can have detrimental effects upon first-pass yields, proportion defectives, repair, and rework in the assembly process. The effect on customer satisfaction with regards to late orders, defective product and returned goods is even more important. It is, thus, important to have a technique to optimize the circuit board's solder pad sizes, and the solder stencil's thickness and aperture sizes so that the correct amount of paste is applied. These techniques involve experience, trial and error, rules of thumb, and the industry's Gurus' textbooks and publications. This paper focuses on solder-joints holding surface mounted terminal blocks in the 20A, 300V, 3.5 and 5 mm pitch size range. INTRODUCTION AND BACKGROUND Printed circuit board (PCB) mounted terminal blocks are comparatively large and are subject to human factors during wiring installation at downstream processors & end users. Most designers are familiar with through-hole wave (THW) mounted terminal blocks, headers, plugs, sockets and pin strips. Wave soldering processes with their virtually unlimited supply of heat & solder metal have proven to the world that they can successfully solder an extensive variety of PCB pin and PCB hole combinations. Genuine surface mount technology (SMT), through-hole reflow (THR) and through-hole wave solder joints are all viable methods of attaching terminal blocks to printed circuit boards (PCB). The SMT components leads are attached to one surface of the PCB on a flat solder land or pad. In both THW & THR, the component leads penetrate the PCB. This paper will focus on pure SMT solder joints that have been processed by the application of paste onto the PCB where the component is picked & placed onto the PCB and the assembly is then passed through a reflow oven.

Soldering technology has undergone recent changes instigated by the European Union’s (EU) imposition of EU directive: Restriction of Hazardous Substances in Electrical and Electronic Equipment (RoHS) Directive 2002/95/EC. The RoHS directive stipulates that lead, mercury, cadmium, hexavalent chromium, polybrominated diphenyl ether and polybrominated biphenyls will be banned from new electrical and electronic equipment as of July 1st, 2006. By eliminating these substances from new equipment the RoHS directive complements the EU’s Waste Electrical and Electronic Equipment (WEEE) Directive 2002/96/EC. The WEEE directive includes substances most of which are not used in electrical connectors. Other nations and private corporations have adopted virtually identical regulations. The net effect on terminal blocks for SMT and PCB applications has been the quasi-elimination of the ubiquitous 90/10 Sn/Pb plating on the PCB leads and all other areas. This plating met some telecom company requirements for the mitigation of tin whiskers and had been actively marketed by the suppliers as being more solderable than pure Sn. The 90/10 Sn/Pb has in most cases been replaced by pure Sn, with the mitigation of tin whiskers being handled by using matte or reflowed versions of this coating with suitable barrier under coatings. The pure Sn coating has the advantage of being backwards compatible. It can be used in RoHS & non-RoHS soldering processes. The eutectic Sn/Pb soldering process is still in commercial use for many reasons. For example military equipment and industrial process control equipment need not comply with RoHS. The Pb free processes are operational in many sectors. Export oriented original equipment manufacturers (OEMs) tend to err on the side of “when in doubt it must be RoHS”. Contract electronic manufacturers (CEMs) are concerned about meeting customers’ requirements. Both are concerned with contamination of one process with the other. When possible and relevant this paper will discuss the nuances of each.

SOLDER PASTE, STENCILS AND PCB PADS Solder paste is a mixture of metals particles, binders and fluxes. It is solid in its container but it flows and liquefies under pressure & agitation; it can be deformed and spread like peanut butter. Solder paste is thus said to have thixotropic properties.

For example, 90 weight % metal in paste : A 63/37 Sn/Pb eutectic paste, 90 wt% metal. Metal vol%=100(0.9/8.4)/(0.9/8.4+(1-0.9/0.8)=46.2%

The binders and fluxes have low melting points and low boiling points. They are volatile and they boil or vaporize away during the reflow process. This occurs at temperatures typically and approximately from 160 to 220 C. This vaporization or outgassing can detrimentally affect solder joint porosity when it occurs when the fused (molten) metal is in its solidification process.

A Sn-3.5 Ag paste, 90 wt% metal. Metal vol%=100(0.9/7.5)/(0.9/7.5+(1-0.9)/0.8)=49.0 %

Conversely to the volatile materials, the metal in the paste stays and forms a solder joint between the PCB solder pad and the components PCB leads. It is important to know the proportion of metal in the paste. This proportion is often referred to as percent metals.

A Sn-3.5 Ag paste, 85 wt% metal. Metal vol%=100(0.85/7.5)/(0.85/7.5+(1-0.85)/0.8)=37.7 %

Paste volume calculated by multiplying the stencil thickness by the area of the stencil apertures. For example, the volume of paste deposited: By a 0.127 mm thick stencil and 2mm x 5 mm aperture. Volume = 0.127 x 2 x 5 = 1.27 mm^3 By a 0.127 mm thick stencil and 3 mm diameter aperture. Volume = 0.127 x 3^2 x π / 4 = 0.90 mm^3 It is important to know the volume of metal deposited on the pad through these apertures because this will have a profound effect on the solder joints shape and appearance. All of which affect the solder joint’s strength and its resulting electrical & thermal conductivities, first pass yield and infield reliability. This proportion of metal is usually expressed as a weight percent (wt%). It is very practical to know the percent metal by volume (vol%). This is calculated by taking into account the specific gravity (SG) of the metal and the binder. Table 1 has some specific gravities of important solder paste constituents. Table 1 Solder paste material specific gravities Material SG g/ml Binder 0.8 63/37 Sn/Pb 8.4 Sn pure 7.3 Pb pure 11.3 Sn-3.5Ag 7.5 Sn-3.5Ag-0.7 Cu 7.5

For example, 85 weight % metal in paste: A 63/37 Sn/Pb eutectic paste, 85 wt% metal. Metal vol%=100(0.85/8.4)/(0.85/8.4+(1-0.85/0.8)=35.1%

Metal volume proportion (vol%) is a function of the weight proportion (wt%) and the specific gravities of the metal and other constituents. One can now calculate metal volume: A 0.127 mm thick stencil and 2mm x 5 mm aperture. A Sn-3.5 Ag paste, 90 wt% metal. Volume metal = volume paste x %metal/100 = 1.27 x 49/100 = 0.62 mm^3 A 0.127 mm thick stencil and 2mm x 5 mm aperture. A 63/37 Sn/Pb eutectic paste, 85 wt% metal. Volume metal = volume paste x %metal/100 = 1.27 x 35.1/100 = 0.45 mm^3 Final metal volume is a function of metal volume proportion and the stencil apertures thickness and area. Final metal solder fillet and joint shapes and appearances will be subsequently affected by: -the solder metal volume -the solder metal type & fluxing -the surface finishes of the component leads & PCB pads. -the area and shape of the solder land, which may be larger or smaller than the stencil aperture -the PCB leads which may be smaller or the same size as the solder lands -excess, insufficient or adequate solder volumes created by all this. If one assumes a 1/1 stencil aperture to solder pad area ratio, flat solder deposition, 90 wt% metal Sn-3.5Ag solder; a 0.127 mm thickness stencil would lay down 0.127 mm thick of paste which would become 0.062 mm of solder metal The authors have developed spread sheet algorithms that calculate solder metal volumes and thickness based upon stencil, pad and solder specifications. Table 2 has calculated the volume and thickness values at aperture/pad ratios of 0.9, 1 and 1.1, metal weight % ranging from 75% to 90%

for Sn-Ag3.5 (also Sn-3.5Ag-0.7Cu) RoHS high temperature solder and 63/37 Sn/Pb eutectic which have metal SG of 7.5 & 8.4 respectively. Table 2. Metal volume and thickness for different pastes at different aperture/pad ratios, SG and % metal.

The physics of wetting involve the wetting angle θ (theta). When θ is acute, relatively small and close to 0º a liquid adheres to a surface. Its atoms and molecules have affinity for and are attracted to the atoms and molecules on that solid’s surface. When θ is obtuse and greater than 90º the liquid has little or no affinity for the solid surface. The solid’s surface atoms and molecules repel the liquid’s atoms and molecules. The liquid’s surface tension prefers to coalesce it into spheres rather than spreading out onto that surface. In all cases the materials seek shapes that yield the lowest energy states taking into account the factors of mass, gravity, density, available liquid and available surfaces. Figure 1 shows the uses of wetting angle θ. In the top diagram, θ appears on molten solder wetting a flat surface similar to an empty PCB solder pad. The middle diagram shows the forces involved. F surface = γs x (solid surface area) + γls x (interfacial area) + γl x (liquid spherical area) γs = γls + γl cosθ The bottom diagram shows the assumptions and general geometry of an ideal spherical cap. Figure 1. The wetting angle θ

SOLDER WETTING This paper will not focus on the reflow process with regards to time-temperature profiles, atmospheres, heating methods nor equipment. The purpose of this paper requires a discussion of the physical principles of molten solder in contact with solderable surfaces. Molten solder will wet a solderable surface just like water with detergent wets dishes and spreads and adheres to them. It will bead on a non-solderable surface just like clean water beads on wax paper or a greasy surface. When water is involved it is said that certain solid surface are hydrophilic (water loving) and that others are hydrophobic (water hating). Chapter 2 Wetting of Surfaces in Soldering in Electronics 2nd edition by R. J. Klein Wassink has excellent information on the theory & principles involved.

The spherical cap concept is not only important as a basic assumption for didactic or research purposes. When this occurs in practice it can cause component leads (and thus components) to shift in undesirable ways. A finite volume of solder put onto a finite sized pad will result in a meniscus that bulges upwards. Controlling this is important.

SOLDER FILLETS AND JOINTS Solder fillets are what can be seen when looking closely at the lead to pad interface. The solder seems to adhere to both surfaces creating an internal radius that can be more or less evident. Cylindrical leads lying on flat pads will have filets with a slightly yet significantly different appearance than a filet where the component lead the touches the pad at a perfect right angle. A pad that is much larger than the lead will have a different fillet than a pad that is only slightly larger than the lead. Excess solder will fatten the fillets making them more massive with a larger inner radius. Insufficient solder will thin them out, giving them what is sometimes called a dry appearance. Solder quantity can be adjusted by the stencil aperture to solder pad area ratio. In figure 2 it can be seen that fillets on a top surface are larger. Gravity helps them be more massive. There is a maximum size that the fillet can attain, the elastica. Conversely fillets on a bottom surface are thinner and lighter because gravity helps draw solder away. Fillet edges, where they appear to blend with the solid surface, can be 1/3 mm, 1/2 mm, 1 mm and up to 5 mm distant from a sharp corner. Figure 2. Solder fillet sizes above and below.

The size of a solder fillet is important in terms of solder consumption. It is also important for the strength of solder joint. In many cases the fillet contact area can involve a substantial proportion of the lead to pad contact area. Also of importance is that a fillet shape and thus volume can be adjusted so as to soak up excess solder. By carefully adjusting solder paste mixture, stencil thickness, stencil aperture, solder pad size and component lead sizes and shapes; optimal solder joints can be designed. The optimal solder joint meets strength, component placement accuracy,

first pass yield and infield reliability requirements economically and consistently. The calculation of the required solder metal to attain a good solder joint can be a first order approximation based on experience. It can be calculated from the shapes, sizes and volumes involved by using simple geometry with reasonable approximations. Calculations that assume standard sized fillets on perimeter and standard thickness of the solder layer between the lead and the pad can be found in texts and handbooks as can precise methods of calculations. The authors have developed spreadsheet algorithms that calculate required solder volumes under different component lead and PCB pad conditions. An example is shown in table 3. Table 3 Solder joint volumes calculated for figure 3.

COMPONENT LEADS AND SOLDER JOINTS Terminal block, pin strip and header component leads are typically made of Cu alloy substrates covered in pure Sn solderable coatings. Brass and bronze are often used Cu alloys. They contain Cu and depending on type will contain specific proportions of Sn, Zn and/or Pb. The pure Sn coating will usually be matte electrolytic, hot-dipped or reflowed. Ni barriers are used when required to stabilize solid-state diffusion of certain metals such as Zn, which can form undesirable Sn-Zn intermetallic compounds in the Sn outer layer. Figures 3 and 4 show a pin strip lead’s solder joint and the pin strip itself. Two types of designs compete in this circular lead nail-head type design. In one type the pins are fixed in the strip, must be very coplanar and must be used on very coplanar PCBs. In the other type the pins can independently move with respect to each other and can thus adapt to noncoplanar PCBs thus eliminating open circuits and increasing first pass yields. In the first case coefficients of thermal expansion (CTE) must closely match that of the FR-4 PCB and the elastic modulus (E) must be such that cyclic thermal forces do not cause excessive fatigue in the solder joints. Solder volume excess may cause the strip to lean in one direction. In the second case the CTE & E do not matter but solder volume excess may affect pin parallelism in random directions. In both cases, pin perpendicularity to PCB and parallelism relative to each other are handled by controlling the solder metal volume and fillet sizes. Notice how the fillet interfaces with the outer radius seen in figure 3.

Gull wing leads flare away from the terminal block. When correctly manufactured they will lie flat on the PCB pad. An advantage is that the leads and thus the solder joints are mostly isolated from screw tightening torques and forces. In order to attain strong peel off strengths the leads must have their heals and toes firmly in contact with the PCB pad and firmly soldered. An elevated heel is especially undesirable. Figure 5 shows a satisfactory solder joint. Screw tightened terminal blocks are exposed to the human factors involved in wire installation and maintenance. Excessive forces occur often and strong PCB to terminal mechanical connections are required in order to attain infield reliability. Figure 6 shows the terminal block for that joint. There are two solderable anchoring elements visible on the right and left ends of the block. These greatly increase the strength of the connection by creating a more balanced mechanical system with more points of attachment placed in areas that distribute loads more effectively. Figure 5. A transverse cross-section of the solder-joint of the block in figure 6. The solder pad is approximately 5 mm long x 2 mm wide. The length is seen here.

Figure. 3 The solder joint on the SMT pin strip header in table 3 and figure 4. The solder pad 2.8 mm in diameter.

Figure 6. The SMT gull wing, screw tightened terminal block seen in figure 5. 5 mm pitch, 300 V, 15 A

Figure 4. The SMT pin strip header seen in figure 3. 5 mm pitch, 300V, 10 A.

Component leads can be directly integrated into terminal bodies. This works well with the adaptable coplanar designs where the components leads are free to move with respect to each other. CTE mismatch and thermal expansion and contraction fatigue induced failures are eliminated. Solder joints open circuits are eliminated and joint strength is high because the leads are always at their best contact with the pads. This helps mitigate the stresses imposed by the torques and forces imposed during wire installation and screw tightening. Figures 7 and 8 show such designs.

utmost importance. The people involved must quickly establish functional communications between each other. The customer/supplier interfaces must be defined. This way the implementation of new technologies can proceed in a pleasant and professional manner that inspires confidence and success. Figure 9 shows some of these technologies. Figure 9. Genuine SMT gull wing terminal blocks, pin strips, headers, depluggable and body bottom lead terminal blocks, 150 t0 300 V, 10 to 15 A, 3.5 to 5 mm pitch

Figure 7. Front view of the solder joints of the block in figure 8. The solder pad is approximately 2 mm wide by 5 mm long. The width is what is seen here.

Suggested reading and references in alphabetical order: Environment-Friendly Electronics: Lead-Free Technology Electrochemical Publications Ltd., 2001 Jennie S. Hwang Figure 8. The SMT terminal block with coplanar adaptable integrated body lead seen in figure 7. 300 V, 15 A.

A Scientific Guide to Surface Mount Technology Electrochemical Publications Ltd., 1988 C. Lea Soldering Handbook for Printed Circuits and Surface Mounting Van Nostrand Reinhold Company, 1986 Howard H. Manko Surface Mount Technology – Principles and Practice Kluwer Academic Publishers, 1997 Ray P. Prasad Soldering in Electronics Electrochemical Publications Ltd., 1989 R. J. Klein Wassink

CONCLUSION The authors have found that the learning curve and infant mortality occurring at new users of these technologies have significantly improved in the recent five or ten years. The quick, accurate, pertinent and effective exchange of information between the user and the manufacturer is of

NIST and NIST related websites contain excellent information on the physical properties of solder, PCB and electronic component materials The authors acknowledge the use of R. J. Klein Wassink’s diagrams in figures 1 and 2. The authors would like to express thanks to Heiner Kammann.