Head in Pillow Defects in BGA Assembly

A head in pillow defect is a ball grid array joint in which the solder ball on the package and the solder paste on the board have each melted and solidified without merging. The two form a shape that looks like a head resting on a pillow, with a boundary between them that is visible in a section and nearly invisible in a radiograph. The joint may pass continuity testing and fail later under thermal cycling or mechanical load.

What Happens at the Interface

The ball on the package and the paste on the board melt at slightly different times and at slightly different places. Where both are liquid at the same moment, surface tension merges them into a single joint. Where one solidifies before the other is fully liquid, the interface is mechanical contact rather than a metallurgical bond.

The result is a joint with a continuous electrical path through the contact area, which is why an in circuit test can pass. The contact area is small, and the boundary has no intermetallic layer, so the joint is far weaker and more resistive than a proper one.

The failure appears as an open after a thermal excursion or after mechanical handling. Because the electrical test passed at build, the fault is often reported as a field failure or as a no fault found, and the assembly is not suspected.

Warpage and Its Timing

The package and the board both warp during reflow, and they warp in different directions and at different moments in the profile. A package that is convex when cold becomes flatter as it heats, and the board beneath it sags as its own temperature rises.

The critical moment is when the profile moves from the soak into the peak. If the package has begun to cool while the board is still at peak, or the reverse, the gap between the ball and the paste opens or closes at exactly the time the materials should be merging.

Warpage is measured as a function of temperature, and the useful figure is the difference between the ball and the paste at the moment of peak. A package specification that gives only the room temperature coplanarity does not describe the behaviour that causes the defect.

Paste Volume and Placement

The paste deposit determines the height of the column between the board and the ball. Too little paste and the two do not touch at the peak of the profile, so the ball rests on a thin layer of paste that has already begun to cure.

Too much paste is not automatically better, because the excess can push the ball upward and create a taller column that is prone to collapse. The target is a deposit that leaves the ball in contact with the paste while both are still liquid.

Deposit consistency matters more than the average. A paste volume that varies across the array means that some joints merge and others do not, which is why the defect often appears on one corner of the package rather than uniformly across it.

Detecting the Defect

A radiograph does not reliably show a head in pillow joint, because the two parts of the joint overlap in the projection and the boundary is a plane rather than a void. A careful inspection with a tilted view can sometimes reveal a change in the silhouette, but it is not a dependable detection method.

The reliable methods are destructive. A cross section shows the boundary clearly, and a dye and pry test opens the joint and stains the unbonded interface. Both are sample based, which limits them to a first article and a periodic audit.

Where a non destructive method is needed, the most sensitive approach is a comparison of the joint resistance or the thermal signature against a known good reference. The change caused by the reduced contact area is small, so the measurement has to be capable of resolving it.

Process Causes

The profile is the first variable. A soak that is too short leaves the assembly uneven in temperature, so the package and the board reach the peak at different times. A ramp that is too fast produces the same result, and a peak that is too low prevents the paste and the ball from becoming fully liquid together.

The atmosphere and the flux chemistry contribute. A nitrogen atmosphere can improve wetting but also changes the timing of the oxide removal, and a flux that has lost activity does not clean the ball surface, so the two liquids meet across a film of oxide rather than merging.

Moisture and contamination belong on the list. A package that has absorbed water releases it during the profile, and the vapour at the interface is enough to prevent coalescence on the joints closest to the die.

The Design Contribution

The design sets the ball diameter, the pad size and the paste aperture, and those three numbers decide how much tolerance the process has. A pad that is much smaller than the ball reduces the contact area, and a paste aperture that is much smaller than the pad reduces the deposit.

The package thickness and its internal structure determine the warpage. A thin package with a large die warps more than a thick one, and the effect is larger on the corners than at the centre, which is why the corner joints fail first and why the substrate design is as much a part of the fix as the reflow profile.

Where a design is known to be marginal, a stiffener, a thicker substrate or a different flow can be chosen before the build rather than after the failures. That decision is cheaper at the start of a project than during production.

Prevention

The profile should be measured at the package and at the board, with thermocouples placed to show the difference between them. Establishing the soak so that both reach the peak together is the single most effective measure.

The paste volume should be measured by inspection and kept within a tight band, and the aperture should be sized to fill the joint rather than to conserve paste. The ball surface condition, the storage of the packages and the age of the paste all belong in the same control chart.

Where the defect has already appeared, the analysis should start with a section that shows the interface and a warpage measurement at temperature. Those two pieces of data separate the profile cause from the design cause, and the corrective action follows from which one is present.

Practical Rules

Measure the profile at both the package and the board, control the paste volume with a measurement, and choose a stack that limits warpage. Treat a passing in circuit test as evidence about continuity, not about the bond.

Record the profile and the paste data with the build records and the defect history, and review the inspection technique and the reflow profile control whenever a boundary is found in a section.

Additional Considerations for This Build

Practical attention to voids 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 voids explicitly, together with the tolerance that applies, removes the assumption and keeps the result predictable from batch to batch.

FAQ

Why does an in circuit test not catch a head in pillow joint? The two solder surfaces still touch, so there is a continuous electrical path. The bond is mechanical rather than metallurgical, and it fails later rather than at test.

Can X ray see a head in pillow defect? Not reliably. The two parts of the joint overlap in the projection and the boundary is a plane rather than a void. A section or a dye and pry test is what shows it.

What is the most effective prevention? A profile that brings the package and the board to the peak together, combined with a paste deposit that keeps the ball in contact with liquid solder at that moment.