Solder Joint Defects: Causes and How the Layout Prevents Them

Most solder joint defects are decided long before the board reaches the reflow oven. A stencil aperture, a pad geometry, a mask clearance or a thermal connection can make a joint that is easy to produce or one that is on the edge of the process window, and the process then has to be run perfectly to compensate. This article looks at the common defects, what causes them, and which of those causes live in the design.

The Defect Families

Defects fall into groups according to what went wrong, and the group usually points at the part of the process to investigate.

Too little solder produces a joint with insufficient fillet, a weak connection and sometimes an intermittent open. The causes are paste volume, aperture design, poor paste release from the stencil, a pad that is too small, or solder that wicked away into a via or a barrel.

Too much solder produces bulging joints that can bridge to a neighbour and that may hide a weaker connection underneath. The causes are an oversized aperture, poor stencil sealing, or a component that floats on an excessive paste deposit.

Bridging is solder connecting two adjacent joints. It happens when there is too much paste, when there is nothing to stop the solder flowing sideways, or when the mask dam between the pads is missing or too narrow.

Tombstoning is a two terminal component standing on one end. It is caused by an imbalance between the two joints: different paste volumes, different pad sizes, different thermal masses, or one end wetting before the other.

Cold joints and incomplete reflow occur when the joint never reached the liquidus temperature, usually because the thermal mass of a plane or a component drew heat away faster than the profile could supply it. This is the defect that a thermal relief is designed to prevent.

Voids and head in pillow are internal defects. Voids are gas trapped inside the joint during reflow. Head in pillow is a ball that sits on the paste without coalescing, and it is the classic hidden failure on a ball grid array.

Cracks appear at the interface between the joint and the pad, and they are usually the result of thermal cycling or mechanical stress rather than a defect of the reflow process itself.

Inspection of solder joints on an assembled PCB under magnification

Where the Layout Decides

Six design choices account for most of the defects that a process engineer cannot fix by adjusting the oven.

Pad geometry. The pad has to match the component termination and the stencil aperture. A pad that is too small limits the joint volume; a pad that is too large lets the component float and shift. Where the pad is larger than the termination, the excess solder can form a ball under the part rather than a fillet at its side.

Mask dam width. A narrow or missing dam between adjacent pads removes the barrier that stops solder from flowing across, which is the direct cause of bridging on fine pitch parts. Our guide to solder mask clearance covers the relationship between the opening, the dam and the process registration.

Thermal connection. A through hole pad connected solidly to an internal plane pulls heat out of the joint during soldering, which produces an incomplete joint or an operator who holds the iron on the pad too long. A thermal relief pattern is the standard solution, and the same reasoning applies to a surface mount pad connected to a large copper area.

Via in pad and via proximity. An untented via close to a pad can wick solder away from the joint during reflow, leaving insufficient solder or a solder ball on the surface. Where a via must be in the pad, it should be filled and plated flat.

Copper balance. A pad with a large copper area attached heats and cools differently from an isolated pad. Two pads of the same component should have similar thermal environments, which is one reason a designer should not add a via or a plane connection to only one side of a passive component.

Component spacing. Parts placed too close together restrict the stencil, interfere with the nozzle and remove the room needed for inspection and rework. The courtyard around a component is not decoration; it is the space the process needs.

Where the Process Decides

The remaining causes belong to the assembly house, and understanding them helps a designer know what can and cannot be fixed by the process.

The stencil determines the paste volume: its thickness, its aperture design and its surface finish decide how much solder is deposited and how much is released. The paste itself has a shelf life, a viscosity and a flux chemistry that affects wetting. The reflow profile determines whether every joint actually reaches temperature, and on a board with mixed thermal mass that is a compromise rather than a single correct answer. The atmosphere matters for some assemblies, where nitrogen reduces oxidation and improves wetting on fine pitch work. And handling matters, since moisture absorbed by components before reflow turns into steam and produces popcorning and voids.

How the Defects Are Found

Each inspection method sees a different subset of the defects, and no single method covers everything.

Visual inspection under magnification finds insufficient and excess solder, bridging, tombstoning, solder balls and obvious displacement. It cannot see under a package, and it cannot judge the internal structure of a joint.

Automated optical inspection does the same job faster and more consistently, using a camera and a programmed model of what each joint should look like. It is good at comparing one joint to thousands of others, and it is poor at judging a joint that is unusual for a good reason.

X-ray inspection sees through the board and finds voids, bridging under a package, missing balls and insufficient solder volume on hidden joints. It shows density rather than shape, so a crack or a head in pillow joint may need an angled view or a cross section to confirm.

Cross sectioning is destructive and slow, and it is the reference method. It is used on first articles and on any confirmed suspect, and it is the only way to see the interface between the solder and the pad on a hidden joint.

Electrical test catches what the optical and X-ray methods miss: a joint that looks correct and does not conduct, or one that conducts at room temperature and opens when warm. That is why continuity test and functional test remain necessary even on a board with thorough optical inspection.

Acceptance Criteria and Rework

What counts as a defect depends on the class of the product. A consumer product and a life supporting medical device have different requirements for fillet size, voiding and acceptable rework, and the standard that defines those requirements should be written into the assembly documentation rather than agreed by eye.

The practical consequence for a designer is that the same joint can be acceptable on one product and not on another, which is why assembly drawings should state the acceptance class rather than leaving it to the assembly house to assume.

Rework is the second half of the story. A joint that fails inspection can often be repaired, and whether it should be is a question of class and cost. A reworked joint is generally weaker than one formed in the original reflow, so a design that reworks well – with accessible pads, generous clearance and a clear legend – reduces the cost of the defects that do occur. Where a defect rate is high, the answer is always to fix the cause rather than to add inspection, because inspection only finds defects that have already been made.

Prevention Checklist

  • Match the pad size to the component termination, and check the paste area against the joint volume the part needs.
  • Keep the mask dam above the process minimum between all adjacent pads, and specify the mask convention explicitly.
  • Give every pad that connects to a plane a thermal relief unless it is a power pad carrying significant current.
  • Keep untented vias away from solder pads, and fill and plate any via that has to sit inside a pad.
  • Keep the thermal environment of a two terminal component symmetric, including any via or plane connection.
  • Respect the courtyard, and leave room for the stencil, the nozzle, the inspection camera and a rework tool.
  • State the acceptance class on the assembly drawing.
  • Ask the assembly house to review the paste aperture design for the specific pad shapes in the design, rather than accepting the default equal-to-pad aperture.

PCB manufacturing process

FAQ

  • What is the most common solder joint defect? Insufficient solder, and bridging on fine pitch parts. Both usually trace back to paste volume and to the mask dam.
  • Can a cold joint be seen visually? A grossly cold joint looks dull and irregular, but a joint that is marginally under temperature can look normal and still be weak.
  • Is voiding always a defect? No. It is normal to some degree, and the acceptable limit is set by the product class and agreed in the assembly specification.
  • What is the difference between head in pillow and an open? Head in pillow has the ball touching the paste without coalescing, so it may pass a cold continuity test and fail after thermal cycling. An open is a complete lack of contact.
  • Why do defects appear only on some panels? Usually because the process is marginal and the panel position, the paste age or the oven load varies. That is a process control problem rather than a design problem.

Summary

Solder joint defects are the visible end of a chain that starts with pad geometry, mask clearance, thermal connections, via placement and component spacing. Design decides what the process window is; the assembly house then has to work inside it. A joint that requires the process to be perfect will eventually fail, and it will fail in the field rather than in the factory.

The prevention is straightforward. Size the pads to the components, keep the mask dam, relieve the pads that would otherwise sink heat, keep vias out of the solder path, balance the thermal environment of two terminal parts, and let the layout respect the courtyard. Then agree the acceptance class and the inspection method with the assembly house, because what counts as a defect is a specification rather than an opinion, and the cheapest quality control is a design that is easy to build.

Leave A Comment