Solder Paste Inspection And Print Quality Control

Solder paste printing is the step that determines most of the defects in a surface mount line, and the deposit is the one thing that can be measured before the components are placed. Solder paste inspection does exactly that: it scans the board after printing and reports the volume, area, height, and position of every deposit, using a three dimensional measurement rather than a picture. The result is a process that can be corrected while the panel is still on the line.

This article explains what the inspection measures, how the limits are set, how the data is used to correct the printer, and where the method has limits of its own.

What The Inspection Measures

A three dimensional system projects a pattern of light onto the board and reconstructs the height of the paste from the distortion of that pattern, which allows the volume of each deposit to be calculated from its height and its footprint. The report lists the volume as a percentage of the aperture volume that the stencil should have deposited, the area, the average height, and the offset of the deposit relative to the pad. Those four numbers describe almost every printing fault.

The faults themselves have characteristic signatures. A deposit with half the expected deposit volume and a full footprint has printed through a clogged aperture. A deposit with the right volume but a shifted centre points to a misalignment of the stencil or a board position error. Bridging appears as a connected region between two apertures, slump appears as a spreading footprint with a reduced height, and solder balling appears as small isolated deposits in the space between pads. Each of those leads to a different correction.

Setting Limits That Mean Something

The classic starting point is a volume window of roughly half to one and a half times the nominal deposit, with a tighter band around the target, but a fixed window taken from a general guideline is a poor substitute for limits that reflect the aperture. A small aperture with a low area ratio transfers less paste even when the process is perfect, so a fine pitch pad may sit legitimately below the nominal volume that a large pad reaches. The limits should be derived from the aperture area ratio and from the process capability measured on the actual product.

Two sets of limits are usually defined. The control limits are the values at which the operator adjusts the process, and the specification limits are the values at which the deposit is declared defective and the board is cleaned. Setting both too tightly produces false calls that send good boards to rework, while setting them loosely lets a real trend pass unnoticed. The trend, rather than the individual deposit, is the more useful signal, and it is usually tracked as a capability index on volume for each aperture size on the panel.

Using The Data To Correct The Printer

The most valuable use of the data is the pattern it reveals. Volume that drifts down across the panel over a shift points to paste drying on the stencil and to the stencil cleaning cycle, while volume that is low in one region and normal elsewhere points to poor board support or to a stencil that is not sitting flat. Bridging confined to one component points to an aperture design or a placement issue rather than to a global printing problem.

The printer parameters then follow from the diagnosis. Squeegee pressure that is too high scoops paste out of the aperture, print speed that is too fast leaves the aperture incompletely filled, and separation speed that is too fast tears the deposit instead of releasing it cleanly. The stencil itself contributes through its thickness, its aperture geometry, and its coating, and the paste contributes through its rheology and its age. Changing one parameter at a time and watching the volume response is faster than changing the printer settings by feel.

What Print Quality Depends On

The aperture area ratio is the geometry that governs transfer, and it is the ratio of the aperture area to the area of the stencil wall that the paste must slide along. Below about two thirds, transfer efficiency falls sharply, and the usual remedies are a thinner stencil, a larger opening, or a paste with finer powder. Rounded aperture corners and a slight flare at the bottom of the opening, which is used in fine pitch work, both reduce the adhesion that holds the paste in the wall.

Support and alignment complete the picture. A panel that is not backed up properly flexes away from the stencil when the squeegee passes, which leaves a wedge shaped deposit, and a stencil that is not aligned to the board moves every deposit by the same offset. The consequences of those errors at reflow are described under component shift during reflow, and the pad geometry that sets the aperture under pad design standards.

Correlating The Print With The Joint

The inspection predicts the joint only indirectly. Volume is a good predictor of opens and of insufficient solder, and position is a good predictor of tombstoning and bridging, but the same volume can produce a sound joint in one profile and a cold joint in another, because the flux has to be activated and the alloy has to coalesce. A deposit that is correct in every dimension can still fail if the profile is wrong or the surface is contaminated.

The way to use the data is to correlate it with the end of line result. When a defect is found after reflow, the print record for that panel is retrieved, and the two are compared, which gradually produces limits that are tied to real failures rather than to a general rule. Where the correlation is poor for a particular component, the cause usually lies downstream of printing, and the inspection is then telling the truth about the paste and nothing about the joint.

Pitfalls And Practical Rules

The first pitfall is measuring everything. A program that inspects every pad at full resolution slows the line and generates data that nobody reads, while a program that samples the critical apertures and reports trends gives more usable information. The second is treating the measurement as an audit rather than as a control: the value of the data comes from the correction that follows it, not from the record that it produces.

The practical rules are to set limits from the area ratio, to track volume by aperture size, to watch the position offset as closely as the volume, and to treat a change in the trend as an event that needs a cause. The layout decisions that make printing easier are described under manufacturable design guidelines.

FAQ

Does a passing print guarantee a good joint? No. The inspection measures paste, not solder. Volume and position predict many defects, but the profile, the surface condition, and the flux also decide whether the deposit forms a sound joint.

Why do fine pitch pads often read below nominal volume? Because the aperture area ratio is low, so less paste transfers from the stencil wall. The limits for those apertures have to be set from the geometry rather than from the nominal deposit.

How often should the stencil be cleaned? Frequently enough to keep the volume trend flat, which is usually determined from the data rather than fixed in advance. A cleaning interval that is too long shows up as a slow decline in volume across a shift.