Components Falling Off Is Not Fixed by Adding Solder: Joint Strength Depends on Area, Structure and Termination Plating — 5 Stages from Fracture Surface to Verification

When a component falls off, the first reaction is often “add more solder”, but more solder is not necessarily stronger. A shear test paper states that joint shear force equals the solder’s own strength times the wetted contact area; in that test, tin-lead joints, though the solder itself is weaker, showed higher shear force than lead-free, which the authors attribute to tin-lead’s larger wetted area. A capacitor maker also writes that too much solder puts extra stress on the component and joint, while too little or uneven solder gives weak bonding, rotation or one end lifting. What really matters is where it broke, whether the wetted area is enough, whether the intermetallic layer has grown too thick, and whether the component termination has a nickel barrier. Terminations without nickel, or with too thin a layer, can leach at high temperature, and one user has had field joint failures from this. On the process side, too much paste, too thick a stencil or poor pad design can also lift components.

This article covers how to find the cause when an SMT component falls off after soldering, or comes off with a push, from joint strength, termination plating and process. BGA ball fracture and component body cracking are out of scope.

• Stage 1 | Look at where it broke: in the solder, in the termination’s silver layer, under the pad, or in the component body.
• Stage 2 | Strength depends on area and structure: shear equals strength times area; too much and too little solder are both bad; an over-thick intermetallic layer turns brittle.
• Stage 3 | Check termination plating: nickel barrier, termination leaching, time at high temperature.
• Stage 4 | It may also be the process: too much paste, too thick a stencil, pad design; nitrogen makes problems show more.
• Stage 5 | How to verify: shear testing, cross-sections, comparison with component specs.

Each stage covers what to do and what public sources say, plus how JIAHON can help where relevant; sources and limitations are in the notes at the end. This is the general case; individual cases vary.

Stage 1: Look at where it broke

What to do:

• Keep the fallen part and the board: do not rush to resolder; keep the component and the pads on the board for inspection (this article’s own summary).
• Find which layer the fracture is in: one shear test saw two failure modes, fracture in the joint or failure of the component itself; most broke at the joint interface (source 3).
• Fracture in the termination’s silver layer: the same test found silver on fracture surfaces with an electron microscope and EDS; the authors read this as partly in the termination metallization (silver layer) and partly in the solder (source 3).
• Four locations, four directions: broken in the solder, look first at wetted area; in the termination metal layer, check termination plating first; pad lifted, check the board and loading first; component body cracked, check loading and the component first (this article’s own summary).

What public sources say: a conference paper by a US state university and a materials company (presented at an electronics industry association conference; no year on the page, about 2006) pushed 1206, 0805, 0603 and 0402 chip resistors with a shear tester. It says shear testing showed only two failure modes, joint fracture and component failure, with most samples breaking at the joint interface; EDS found silver on fracture surfaces, which the authors read as partly in the termination metallization (silver layer) and partly in the solder bulk.

How JIAHON can help: the broken component and pads can be photographed with a 7-inch digital microscope (the product page lists 1–1,200× continuous zoom, an adjustable lens angle to avoid glare, and photo and video capture).

Stage 2: Strength depends on area and structure

What to do:

• Shear = strength × area: joint shear force equals the solder’s own shear strength times the contact (wetted) area (source 3).
• More solder is not the same as stronger: in that test, tin-lead solder itself was weaker than lead-free, yet its joints showed higher shear force; the authors attribute this to better wetting and larger contact area (source 3).
• Too much solder is also bad: too much solder causes thermal and tensile stress on the component, which can crack the component or the joint; too little or uneven solder gives weak bonding, rotation or one end lifting (source 1).
• Solder volume depends on the pads: volume depends on the process and pad size; both pads should be the same size, so unbalanced surface tension does not pull the component askew (source 1).
• A thick intermetallic layer turns brittle: in an aging test, the intermetallic layer thickened with aging time and joint strength fell because of its brittleness; as-reflowed joints had higher shear force than aged ones (source 4).
• Reflow profile effects are inconsistent: the effect of peak temperature and time above liquidus on shear force was not consistent across component sizes and pastes (source 3).

What public sources say: the university and materials company paper says shear force equals joint shear strength times contact (wetted) area. Citing a national standards institute’s solder property database, tin-lead solder’s shear strength is lower than SAC, so the authors conclude the only likely reason for higher tin-lead joint shear is larger contact area, supported by optical inspection of fracture areas and consistent with known wetting differences. A capacitor maker’s application notes say solder volume affects reliability; too much causes thermal and tensile stress that can crack the chip or joint; too little or uneven solder causes weak adhesion, rotation or one end lifting; volume depends on process and pad size, and both pads should be the same size. A UK university paper says intermetallic thickness grows with aging, and joint strength falls because of its brittleness; as-reflowed 1206 chip resistor joints had higher shear force than aged ones.

Stage 3: Check the termination plating

What to do:

• Nickel barrier: early component terminations were silver-palladium paste, which solder dissolved during soldering; a nickel barrier was later plated over the silver, and since nickel dissolves very slowly, the problem almost disappeared (source 2).
• No nickel or too thin: some suppliers did not plate nickel, or plated too thin, and terminations leached at high temperature; this recently caused field joint failures for one user (source 2).
• Capacitor maker’s advice: nickel-barrier terminations are recommended for any soldered application; silver-palladium or gold terminations are for conductive adhesive or wire bonding (source 1).
• Keep high-temperature time short: keep time above 230°C during soldering as short as possible, so the barrier does not thin and the bond fail; that capacitor maker offers reinforced terminations for long soldering or multiple reflows (source 1).
• Silver in the solder: some processes use tin-lead solder with 2% silver; silver already in the solder slows dissolution (source 2).

What public sources say: a specialist at a UK national measurement laboratory, writing in an electronics assembly magazine’s defects column (November 2014), says termination leaching was a problem when surface mount was introduced; nickel barriers plated over silver-palladium terminations dissolve much more slowly, so leaching almost ceased to be a problem, and solderability life was extended. However, some suppliers did not plate nickel or plated it too thin, leading to termination loss at high temperature, recently causing one user field joint failures. Surface mount and hybrid users have specified tin-lead with 2% silver, still used by many companies, and the silver slows dissolution. The capacitor maker’s notes recommend nickel-barrier terminations for any soldered application, silver-palladium or gold only for conductive adhesive or wire bonding, and keeping time above 230°C as short as possible to limit barrier thinning and bond failure.

Stage 4: It may also be the process

What to do:

• Common causes of lifting: too much printed paste, too thick a stencil, poor pad design. Nitrogen or vapor-phase soldering wets more quickly, which makes lifting more likely, but nitrogen itself is not the cause (source 5).
• Misprinted paste can also cause drops: with too little or offset paste on a pad, one side takes less solder and is easily pulled up or pushed off (this article’s own summary).
• Placement position: an offset component takes uneven solder at its two ends and tends to rotate or lift in reflow (this article’s own summary, based on source 1’s equal-pad rule).
• Check one at a time: paste printing, stencil, reflow, PCB design and placement can each be the cause; rule them out one by one (this article’s own summary).

What public sources say: a process engineering consultant, in an electronics assembly magazine’s defects column (July 2020), says many papers cover component lifting in reflow; in their experience the common causes are too much paste, too thick a stencil or poor pad design. Nitrogen or vapor-phase soldering wets more quickly and magnifies these problems, but is not itself the cause, and older AOI may miss slight lifting. The capacitor maker’s notes say both pads should be the same size to avoid uneven solder and unbalanced surface tension pulling the component askew.

How JIAHON can help:

• Drops or open joints can come from paste printing, stencil design, the reflow oven, PCB design or placement; JIAHON’s engineering team can help look for the cause across these.
• Paste printing: JIAHON’s article How to troubleshoot solder paste print defects says worn or loose clamp plates let the PCB shift during the squeegee stroke, causing print offset, open joints and similar defects.
• Stencil cleaning: JIAHON’s article Is your stencil cleaning done right? says many people take cleaning to mean “looks clean”, which is not enough for fine and ultra-fine pitch products.
• Placement: JIAHON’s article Notes on learning SMT mounter programming says the machine uses the head-mounted mark camera to read the board’s fiducial marks, compares actual positions with program coordinates, and corrects all placement positions together.

Stage 5: How to verify

What to do:

• Shear test: push the component with a shear tester and record peak shear force and where it broke (source 3); test enough parts under each condition (this article’s own summary).
• Cross-section the intermetallic layer: a cross-section can measure intermetallic thickness, which can be used to estimate joint shear force (source 4).
• Compare with component specs: check whether the termination has a nickel barrier, the recommended soldering conditions and time above 230°C (source 1).
• Compare with process records: look at print volume, stencil thickness, reflow profile and placement offset together with where parts fell off (this article’s own summary).
• Do not just add solder: set solder volume by pad and process, and confirm where it broke before adding solder (this article’s own summary, based on sources 1 and 3).

What public sources say: the university and materials company paper tested 6 parts per component type per board on 3 boards, recording peak shear force and failure mode, and examined fracture surfaces with SEM and EDS. The UK university paper studied joints after reflow and aging by cross-section, measured shear force with a shear tester, and states intermetallic thickness can be used to predict component-to-pad joint shear force.

How JIAHON can help: cross-sections can be observed and photographed with a 4K 3D digital microscope (the product page lists a built-in measurement system and multi-angle rotating display, for joint inspection and defect identification). Contact JIAHON or email [email protected].

Notes: sources and limitations

• Translation: the sources are in English; statements here are paraphrased, not quoted.
• This article’s own work: items marked “this article’s own summary” are practice notes, not source statements.

This is a general overview, not a failure-analysis report, and not an endorsement of any component maker, supplier or service provider. Go by the component maker’s data sheet, your own test results and your own evaluation. This is the general case; individual cases vary.