Head in Pillow and Solder Ball Joint Defects

A head in pillow defect is a joint where the ball and the paste have both melted but have not united. The ball keeps its spherical shape, the paste has formed a fillet below it, and the two meet along a boundary that carries little current and no mechanical strength.

The joint passes an X-ray view that only checks for missing balls, and it passes a functional test at room temperature. It fails later, often after a thermal cycle, which is what makes the defect expensive.

What Happens at the Interface

The ball and the paste are separate sources of alloy that must merge while both are liquid. If the paste solidifies before the ball has become fully liquid, or if an oxide layer separates them, the two never coalesce and the boundary is fixed.

The result is a joint with a visible gap at the interface when it is sectioned, and a resistance that is higher than a sound joint but not high enough to fail a test. The defect is a process window problem rather than a defect in either material.

Warpage and Separation

The most common cause is dynamic warpage of the package or the board during reflow. As the assembly heats, the package bows and the ball lifts away from the paste at the centre of the array, then returns as the board cools, leaving the two surfaces joined only where they happened to touch while liquid.

Warpage depends on the package construction, the board thickness and the copper balance, and it changes with the temperature. A package that is flat at room temperature is not necessarily flat at the peak of the profile.

Profile Effects

A soak that is too short leaves the package body lagging behind the board, so the ball is still heating while the paste is already melting. A peak that is too low prevents full melting of the larger balls in the array, which is a particular risk when the package uses a higher melting alloy.

The profile should be measured with a thermocouple under the package as well as on the board surface, so that the actual difference between the two temperatures is known. This is the same requirement that governs the zone verification practice for any thermal mass.

Oxide and Surface Condition

An oxidised ball surface does not wet, and the oxide can form during storage or during the early part of the profile before the flux has acted. Humid storage and a long floor life both increase the oxide thickness on a solder ball.

The flux in the paste must be able to reduce that oxide at the temperature where the two materials meet. Where the paste flux is weak or the oxide is heavy, the joint forms mechanically but not metallurgically.

Paste Volume and Placement

The paste deposit has to be enough to contact the ball and form a fillet, and not so much that it lifts the package. A deposit that is too thin cannot bridge the gap once the ball has settled, which produces an open rather than a head in pillow defect.

Placement force and nozzle condition influence how far the ball is pressed into the paste, and a package placed with too little force may rest on the highest deposit rather than settling onto all of them.

Detection

Transmission X-ray shows the ball as a circle and the paste as a fillet, and the interface between them appears as a lighter ring where the material is thin. The interpretation requires the operator to look for a specific feature rather than for a missing part.

Computed tomography resolves the interface more clearly, and sectioning a sample remains the definitive check. Where a product has a history of the defect, a slice is worth the cost of a destroyed assembly.

Material Factors

The ball alloy and the paste alloy should be compatible in melting range, and a wide gap between them makes the process window narrow. A package supplied with a high melting ball is more demanding than one supplied with the same alloy as the paste.

The ball diameter also matters, because a larger ball needs more heat and takes longer to become fully liquid. Arrays with mixed ball sizes are the hardest case and need a profile that satisfies the largest one.

Board and Package Design

A board that is thin and flexible warps more than a thick one, so a package on a thin board is more exposed. Copper balance across the array, a stiffener under the package and a symmetrical stack up all reduce the movement.

The package supplier’s recommended land pattern and paste aperture are developed to work with the warpage of that package, so departing from them without a reason removes a known good starting point.

Correcting the Process

The first change is usually an extension of the soak, which lets the package catch up with the board before the alloy melts. The second is a slightly higher peak, where the components and the laminate allow it.

If the defects persist, the paste aperture is increased within the limit that keeps the standoff correct, and the placement force is verified. Each change should be made one at a time so that the effect can be measured.

Verification and Records

Verification uses the same methods that detected the defect, applied to a sample after each change. The profile, the paste lot and the package lot should be recorded with the result, because the defect is sensitive to all three.

The records sit alongside the process data described for manufacturing processes, and they are what allows a supplier dispute to be settled with evidence rather than with opinion.

Additional Considerations for This Build

Practical attention to reflow profile pays for itself here, because it decides whether the finished board behaves as the drawing intended. Where the requirement is not stated on the fabrication drawing or in the assembly notes, the shop has to assume a default, and that default is rarely the value the design was simulated with. Stating reflow profile explicitly, together with the tolerance that applies, removes the assumption and keeps the result predictable from batch to batch.

Process Control and Verification

On a design of this kind, head in pillow is the item that decides how the rest of the board is arranged. Reviewing the design before the data is released is cheaper than correcting it after the panel is in the tank, because every step downstream inherits the decision made at the front end. Documenting the assumption is part of the design work, and a short note on the drawing prevents a question that would otherwise arrive a day later and cost a day of schedule.

The process window is set by the narrowest step in the flow, so an improvement anywhere else shows up as margin rather than as yield until that step is addressed.

Related reading: our fabrication notes, board quality and design release notes cover the same ground.

A first article check confirms that the process and the drawing agree on the points listed above, and that the coupon data supports the values used in the design.

Where a measurement falls outside the expected window, the sample is retained so that the cause can be established before the balance of the batch is released.

A first article check confirms that the process and the drawing agree on the points listed above, and that the coupon data supports the values used in the design.

Where a measurement falls outside the expected window, the sample is retained so that the cause can be established before the balance of the batch is released.

FAQ

Can head in pillow be detected electrically? Sometimes, as a higher resistance on the affected net, but a functional test at room temperature frequently passes.

Does a nitrogen atmosphere help? It reduces oxidation and widens the window, and it does not correct a warpage problem.

Is the defect caused by the paste or the package? Usually by the interaction of warpage and the profile, which is why the correction is in the process rather than in either material.

How is a rejected assembly repaired? By removal and replacement, since reflowing the incomplete joint again rarely produces full coalescence.