Introduction
This information is meant to elucidate situations where solderjoints are not completely fulfilling the visual requirements of the perfect solderjoint, but are still, without additional touch-up, good soldered joints. This means that they perform both the mechanical and the electrical functions without failures. Additional touch upon these joints will not improve their functional quality and is therefore not advised.
Visual aspects of good soldered joints
Visual aspects of good soldered joints are:
• Good wetting of the surfaces, solder should have flowed evenly over the soldered surfaces.
• Sufficient amount of solder, the solder should provide a reliable mechanical and electrical contact.
• Uninterrupted and smooth surface*), however small craters in PTH's do not need touch-up.
*) In lead-free soldering the fillets have mostly a rough surface or even small cracks. These are effects that are related to the SAC-alloys used in lead-free soldering.
These items will be valid as a general basis for the inspection of solderjoints, also where we discuss the deviations from the perfect solderjoint, these requirements have to be fulfilled.
Rough joint surfaces
In lead-free solder alloys containing tin, silver and copper are mostly used, so-called SAC-alloys.
The combination of tin-silver and tin-copper and tin-silver-copper have each their own eutectic composition.
As a result of that, small volumes of such compositions can precipitate out from the ally during solidification of the solder in the joint. Since these compositions have a different melting point, during solidification of the solder on the joint, parts can already be solid while the remaining solder in the joint is still in a liquid phase. Since the solder also shrinks about 4% during this process, the shrink will only take place at that part of the alloy mix that has the lowest melting point. This will finally cause the rough surface, because the solidified parts do not take part of this volume shrink, so that these crystal profiles can be seen at the surface of the joints.
• A rough solder fillet surface in lead-free soldering is an effect that is part of the process and can not be eliminated with the common SAC-alloys.
Cracks in lead-free solderjoints
The movement during the soldering process of the solder pads that are connected to the metal (copper) barrel causes cracks in the solder fillet. The difference in thermal expansion between the base material (FR4-epoxy) and the copper barrel cause this movement. Due to the greater expansion of the base material the solderpad(s) get a wedged shape during the soldering process. As the solder starts to solidify, the base material returns to its original shape. It is this movement that at that moment creates stress in the weak solder fillet. As a result of that ruptures or cracks may show up in the solder fillet.
• A crack in a lead-free solder fillet is an effect that is part of the process and can not be eliminated with the common board materials.
Craters or blowholes in solderjoints
A crater or blowhole in a plated through hole (PTH) solderjoint can be formed, if during the soldering process absorbed humidity is outgassing from the board material and this outgassing can escape via gaps in the PTH-metallization. This means that this phenomena will only occur if the metallization shows some cracks, because it is impossible for a gas to pass a closed metal barrel. As long as there is evidence of good solder wicking in the holes small defects in the hole metallization have no major effect on the joint reliability as such, but it will create craters or blowholes in solderjoints.
The reason for crater formation is that the solidifying solder will still generate heat in the board material, so that the outgassing still continuous. At the same time the solder starts to solidify from the topside of the board. This means that the gas can now only escape through the solder which is still liquid, which is the solder at the underside of the joint. As the joint solidifies further the crater will be formed. As long as the crater area on the surface is less than 1/4 of the joint circumference, touch-up is not necessary, since it is a reliable joint.
• A crater in a plated through hole less than 1/4 of the joint circumference will still guarantee a good soldered joint and needs no touch-up.
Lifted components
For leaded components mounted in holes, there is the requirement that the protruding length on the solder side should be at least 1 mm, in order to give the joint sufficient strength in case of single sided boards with no plated through holes. For boards with PTH's, this requirement is not directly necessary since the solder in the hole will give the joint sufficient strength.
If there is evidence that the solder has wicked-up to the lead on the component side (topside) of the board, there is no need in this case for the leads to protrude to the solder side, because the wicking to the lead on the topside is only possible if a sufficient length of the lead is in the hole.
If the lead length in the hole is too short, it is impossible for the small amount of solder in the hole during the solder process, to heat up the lead sufficiently for good wetting. In that case the solder will never wick-up to the lead.
When inspection on the topside is impossible, one must in such cases have evidence that the lead contour is visible in the joint on the solder side.
• If one can inspect soldered joints on the topside for good wetting, there is from the solder quality point of view, no necessity to correct lifted components, as long as this does not interfere with other requirements.
Fat solderjoints
In wave soldering with boards having no solder resist, one will never find fat joints, because due to the surface tension of the liquid solder the excess solder, that can remain on a joint during the separation of the board from the solderwave, will be pressed to the tracks that are connected to the joint. Excess solder that remains on a joint where the tracks are covered with solder resist will stay on that joint, because there is no way for the solder to drain off, since the tracks are covered with solder resist. Such a fat joint is however a perfect joint, since the solder in wave soldering will only adhere to a solder connection when the solderability of this connection is perfect. In the case that the solderability of a joint is not perfect, than in wave soldering such a joint will show non- or imperfect wetting and will never give a fat joint. So a fat joint in wave soldering is always a perfect solderable joint. The reason that it is fat, is due to the fact that during the separation of the solder from the last solderwave more solder was left on that particular joint. Such a joint should not be touched-up, since it is a perfect functional joint and we know the reason why these joints can be formed. It is impossible in a wave soldering process on boards covered with solder resist to avoid such joints, because the separation of the solder from the board during the departure from the solder wave is very much depending on the lay-out.
Further on, the solder flow on the front end and the tail end of a SMD-component placed in the solder direction is not the same, due to the flow behaviour of the solder wave. Fat solderjoints are almost only found on the tail end of such a component. To prevent fat joints and solderbridging, "solder thieves" are placed at SO-IC's used in wave soldering.
• Fat solderjoints in wave soldering on boards with solder resist are good joints and should not be touched-up.
Lean solderjoints
Lean solderjoints on SMD's will be formed due to the small dimensions of the solder pad in relation to the lead dimensions. This is often the case with SO-IC lead connections. The solder in the capillary between the lead and the pad is sufficient for a reliable contact. Control of good wetting can be found at the small contour of the solder between the lead and the pad. If there is no interruption in this contour there is evidence of good wetting.
Lean solderjoints can also be formed at the SMD-components that are placed on so called "via's in pad's". If the via hole is not completely filled during the contact with the chipwave, the solder may be wicked from the solderpad after the contact with the second wave. The reason is, that after the separation of the board from the second wave the inside board temperature is still increasing for a while, due to the solder which is in the holes. As a result of that, the solder in the via hole will still rise in the hole by capillary action. Since there is no contact with the wave anymore, the solder that goes in the via hole is "stolen" from the joint which is on that via pad. Due to the capillary between the SMD-metallization on the underside of the SMD and the pad on which it is placed, there is still a sound solderjoint if there is no interruption in the contour between pad and SMD-metallization and the solder meniscus has wicked-up to at least 1/3 of the height of the metallized surface.
• Lean solderjoints on SMD's in wave soldering are good joints if they show evidence of good wetting between SMD-metallization and solderpad and the solder meniscus has wicked-up to at least 1/3 of the height of the metallized surface.
Via holes
Via holes are holes for electrical connections between conductive patterns that lay on different surfaces/levels of the printed circuit board. These holes are not meant to mount leaded components. They provide their electrical function by the metallization of the hole wall. There is therefore no need for the solder to fill these via holes, unless it is necessary for reasons of electrical testing.
• Via holes do not need to be filled with solder.
Misplacement of SMD-components:
Misplacement of SMD-components, such as longitudinal-, transversal- or rotational-shifting, is allowed from the soldering point of view, as long as the solder joint contact between the solderpad and the metallization of the SMD component is at least provided over half the width of the SMD-metallization. A second demand is, that the soldered part should be in contact with the solder pad without airgap, so that the capillary between the metallization and the solderpad is filled with solder, or the solder meniscus has wicked-up to at least 1/3 of the height of the metallized surface, or to a wicking height of 0,4 mm. A larger height is allowed but is not necessary.
• Misplacement up to half the metallization width of the SMD's will still give good solder joints, provided there is evidence of good wetting.
Solderballs
To understand the formation of small solderballs on the solder side of the board, one has to look to the energies that are acting upon the solder and the board. Due to adhesive and capillary force the solder will stay on solderable surfaces, such as pads and joints, forming the solder connections. The cohesion forces will give the solder connections its final form, the cohesion is always less than the adhesion. The adhesive/cohesive force balance is effected by the surface tension of the liquid solder.
Next to the solderable parts the solder comes in contact with the board material more in particular the solder resist, here the interfacial tensions between the solder and the resist are of importance.
These interfacial tensions are acting opposite the adhesion- and cohesion forces. The "value" of these interfacial tensions between the solder and the solder resist is of direct influence on the amount of the small solderballs.
Systematic solderballs are formed at the point were the board leaves the solder wave and the remaining solder between the joints is not totally wicked to the joints (adhesive forces), or the solder is not flowing back in the wave (cohesive forces).
Larger interfacial tensions between the solder and the board will give more systematic solderballs.
The adhesion between the solderball and the solder resist is in all cases very strong. The solder-balls could only be removed by disrupting the interface between the solder and the solder resist. Under sufficient magnification often one can find a "crater" at the point were the solderball was adhering on the solder resist, or one can find resist particles on the solderball that was removed. As a result of this one should consider the necessity for the removing of small solderballs from a soldered board. At least the solderballs on the solder side can remain if they are smaller than 0,15 mm, in which case they normally do not interfere with electrical restrictions such as minimum clearance distances.
• Due to their strong adhesion small solderballs should not be removed if they do not interfere with electrical restrictions, such as minimum clearance distances.
Summary of conclusions
• A rough solder fillet surface in lead-free soldering is an effect that is part of the process and can not be eliminated with the common SAC-alloys.
• A crack in a lead-free solder fillet is an effect that is part of the process and can not be eliminated with the common board materials.
• A crater less than 1/4 of the joint circumference will still ensure a good soldered joint and needs no touch-up.
• If one can inspect soldered joints on the topside for good wetting, there is from the solder quality point of view, no necessity to correct lifted components, as long as this does not interfere with other requirements.
• Fat solderjoints in wave soldering on boards with solder resist are good joints and should not be touched-up.
• Lean solderjoints on SMD's in wave soldering are good joints if they show evidence of good wetting between SMD-metallization and solderpad.
• Via holes do not need to be filled with solder.
• Misplacement up to half the metallization width of the SMD's will still give good solder joints, provided there is evidence of good wetting.
• Due to their strong adhesion small solderballs should not be removed if they do not interfere with electrical restrictions, such as minimum clearance distances.
ITW EAE Support Center