Solder Balling in Reflow: Causes, Detection and Process Fixes
Small spheres of alloy sitting beside a joint, or in a scatter across the mask between pads, are one of the most common reflow findings and one of the least well explained. Solder balling has several distinct causes, and they are separated by where the balls sit and by what the profile was doing when they formed.
What Solder Balling Looks Like
The balls are typically between 25 and 150 micrometres across, which puts them below the resolution of an unaided eye and above the threshold that a functional test will notice. They appear as a residue beside a joint, as a ring around a pad, or as a scatter over an area of mask where solder paste was never printed.
Two patterns are worth separating at the start: balls attached to the fillet or sitting inside the joint area, and balls isolated on the mask away from any paste. The first points to the paste and the print, while the second usually points to flux behaviour or to contamination on the board.
Where the Balls Come From
The alloy in a ball has to come from somewhere, and the sources are the paste itself, the powder that has oxidised and separated, spatter thrown from the joint during the ramp, and alloy that was already on the board from the finish. Each source leaves its own signature in the size and the distribution of the balls.
Flux spitting is the mechanism that surprises people most, because the flux boils and ejects droplets of alloy or of paste that then land on the mask. The droplets come from the deposit, not from the joint after it has formed, which is why the balls are found away from the pad rather than on it. The size distribution is a second clue, because a scatter that is uniform in size came from a single event, while a mixture of sizes usually means two mechanisms have been running at once.
Paste Slump and Its Role
If the printed deposit slumps before the alloy melts, part of it can reach the edge of the pad and split off as the paste dries and reflows. Paste slump depends on the metal content, the viscosity and the moisture the paste has taken up since the jar was opened.
Paste at the end of its open time slumps more, because the solvent has evaporated and the powder has begun to separate. A balling problem that appears late in a shift, or on the boards printed last, is a paste handling problem before it is a profile problem, and the print parameters are the first thing to check. Where the paste is dispensed rather than printed, slump matters less because there is no aperture to release from, but the shape of the deposit still decides which way the solvent escapes.
Preheat Ramp and Flux Volatiles
The preheat ramp decides how the solvent and the volatile components of the flux leave the deposit. A ramp that is too fast boils the solvent under a skin that has already formed, and the resulting bubble throws material out of the deposit; a ramp that is too slow dries the paste so thoroughly that the flux cannot do its job.
The usual band is 1 to 2 degrees Celsius per second up to the soak, which lets the volatiles escape without bubbling. The measurement is taken on the board with a thermocouple, and the profile is checked after any change to the paste or to the board’s thermal mass, since both change the ramp the paste actually sees.
Flux Chemistry and Spitting
The quantity of flux in the deposit and its solvent system set the risk of flux spitting. A paste with a high solvent content is more prone to it, and the effect is worse where the deposit is thick, because the solvent has further to travel before it reaches the surface.
Lowering the deposit volume or moving to a paste with a different solvent system both reduce spitting, and both are changes that can be tested on a small build. The flux solids content is the figure to compare between two pastes, since a high solids flux leaves less to boil off. Ambient humidity in the printing room changes how quickly the deposit takes up moisture from the air, which is one reason the same paste and the same profile can behave differently in summer and in winter.
Stencil, Aperture and Print Quality
Paste that is smeared on the mask beside a pad has already left the aperture and will reflow into balls wherever it landed. That smear comes from a stencil that is not gasketing properly, from an aperture that is too large for the pad, or from a print that happens at the wrong speed for the paste.
The aperture is normally sized to the pad rather than to the joint requirement, and the two can conflict. Where the aperture is larger than the pad, the excess paste has nowhere to go and spreads onto the mask, which our paste volume notes describe from the deposit side.
Surface Finish and Mask Effects
The surface finish on the pad and the condition of the mask both change how the paste behaves. A rough hot air levelled finish holds paste differently from a smooth organic coating, and a mask with a high surface energy encourages the paste to wet the mask surface rather than stay on the pad.
Mask that has not been fully cured can also outgas during reflow, and the escaping vapour throws alloy particles away from the joint. Our surface finish notes compare the finishes on the same basis, and the comparison is worth making before the paste is blamed.
What Counts as a Defect
The acceptance rules distinguish between balls that are entrapped in the joint or that reduce the spacing between conductors, and balls that sit on the mask away from any conductor. The first group is a defect, while the second is often acceptable when it is limited in number and in position.
The judgement is made against the standard rather than against the customer’s reaction to a photograph, and our board failure notes describe how the different findings are catalogued so that a trend can be seen rather than a single board argued about.
Isolating the Cause and Recording the Fix
Balling is investigated one variable at a time, because the causes overlap: a paste change, a profile change, a stencil change and a board finish change can produce the same appearance. A designed comparison on a small number of boards gives an answer in a day, while a change made on a hunch gives an answer that cannot be repeated.
The record should carry the paste lot and open time, the print parameters, the profile with the measured ramp, the board finish and the inspection result. Our profile verification notes describe how the ramp is measured, which is the figure that most often explains a change in balling between two otherwise identical builds.
FAQ
Are solder balls always a defect? No. Balls isolated on the mask, away from conductors and in limited numbers, are commonly accepted. Balls that are entrapped in a joint or that reduce spacing are defects, and the distinction is written into the acceptance standard rather than decided per board.
Why did balling appear after a paste change? Because the new paste has a different solvent system or a different metal content, and both change how the deposit dries and how much material is available to be thrown out. The profile usually needs a small adjustment to match the new paste.
Can the profile be changed to fix balling without touching the paste? Sometimes, and lengthening the preheat is the usual first move. The change is measured on the board rather than in the oven air, because the deposit’s own temperature history is what decides whether the solvent boils or evaporates.
