Print Offset Correction: Stencil Alignment and Paste Registration

A print offset of a tenth of a millimetre is invisible on a coarse board and fatal on a 0.4 mm pitch device, because the same displacement that leaves a comfortable deposit on one pad spills onto the mask beside the next. Correcting it is a measurement exercise: find the direction, find whether it is global or local, and fix the cause rather than the number on the screen.

Where Print Offset Comes From

Offset is the difference between where the aperture was drawn and where the paste lands relative to the pad. It accumulates from board clamping and support, from the stencil’s position in its frame, from the tension of the foil, and from the printer’s own vision and motion. A small amount is present in every process; the question is whether it stays inside the fraction of pad width that the design can tolerate.

Because several of those terms change during a shift, offset behaves like a slow drift rather than a fixed error. Board support settles, the stencil warms, and the paste reservoir changes, so a correction made at the start of a run may no longer be right two hours later. Treating offset as a static number is the reason it is often over-corrected.

Measuring Offset on the Board

The most direct measurement is the deposit position on the pad, taken from the <a href="https://www.gopcba.com/pcb-spi-guide/” title=”SPI”>SPI if the line has one, or from a sample board under a measuring microscope if it does not. Measure the deposit centroid against the pad centroid in both axes, and record the value with a sign so the direction is unambiguous.

As a rule of thumb, an offset above about 25 percent of the pad width begins to starve one side of the joint and overload the other, and above 50 percent the deposit touches the mask and can bridge. On a 0.2 mm pad, 25 percent is 0.05 mm, which explains why fine pitch demands a correction capability that coarse boards never need.

Using Fiducials Correctly

Machine alignment depends on the fiducial targets, and the vision system is only as good as the contrast and geometry it is given. Fiducials should be the specified diameter with a clear mask border, and the pair or trio should span the board so that rotation and scale errors are separated from translation. Two targets close together cannot resolve rotation.

A fiducial that is partially covered by mask, or one sitting on a copper region of similar reflectivity, produces a centroid error that the machine faithfully transfers into the print. When offset is large and consistent in a rotational direction, check the fiducials before adjusting anything on the print head.

Machine Correction Versus Process Correction

The printer offers an X and Y offset correction, and it should be used as a fine adjustment after the mechanical causes are addressed, not as the first response. Entering a large offset to compensate for a misaligned stencil leaves the machine outside its normal working range and hides the underlying problem until the next product runs on the same line.

A practical procedure is to fix the mechanical alignment, print a test panel, measure, and then apply whatever small correction remains. Record both the measured offset and the correction applied, so the difference between them is visible as the process contribution rather than being absorbed into a single number.

Direction and Its Causes

The direction of the offset points to the cause. A uniform shift in one axis suggests a stencil or fixture position error. An offset that grows from one side of the panel to the other suggests a scale or rotation error, which usually comes from the fiducial system or from panel shrink. An offset that changes with separation speed suggests paste behaviour rather than geometry.

Diagnosing the direction rather than the magnitude saves time. Two offsets of the same size can require completely different fixes, and the axis, the sign and the distribution across the panel are what separate them.

Global Versus Local Offset

Global offset applies across the whole panel and is corrected by a single adjustment. Local offset affects one area: a region where the board is unsupported, a corner affected by clamping, or a spot where the stencil is not in contact with the board. Local offset cannot be corrected globally, and attempting to do so moves the rest of the panel out of specification.

Local variation is usually a support or flatness issue. Check the board support pins, the vacuum grooves and the stencil contact pattern, then re-measure in the same location. Where a local offset follows a particular product, the fix belongs in the tooling rather than in the offset value.

Offset on Panel Versus Board

Printing a panel means the offset has to be acceptable for every board on that panel, and the position of the board within the panel matters. A panel that grows during processing moves the outer boards relative to the fiducials, so the middle boards may be centred while the edges are not. Measuring only the first board on a panel hides this pattern entirely.

Where the panel carries different products or different pitches, the offset has to satisfy the tightest one. That requirement should be reflected in the panel layout, because a design that places a fine-pitch device at the panel edge is asking the process for capability it does not need to provide elsewhere.

Verifying Correction Over a Run

After any correction, print a series of boards and measure deposit offset on each rather than on the first one only. A correction that looks right on board one and drifts by board twenty has been applied to the wrong cause, and the run is at risk for the rest of the shift. Track the offset value across the run so the direction and rate of drift are visible.

Where the line runs multiple products, verify the correction again after each changeover, since the stencil, support and fiducial positions all change with the product. A correction is a property of a setup, not of a machine.

Records and Drift Detection

Record the measured offset per product per shift, with the axis and the correction applied. Over a few weeks the record shows whether the process is stable or whether a mechanical element is wearing, and it turns an intermittent print complaint into a trend with a start date.

Combine the offset record with the deposit volume record from the same boards, since volume and offset often fail together when the stencil is not seated correctly. Two measurements from one board cost little and separate a printing problem from a paste problem.

Additional Considerations for This Build

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

FAQ

How much print offset is acceptable? Keep it under about 25 percent of the pad width. On a 0.2 mm pad that is 0.05 mm, which is why fine pitch drives the alignment capability requirement.

Why does the machine offset correction not fix my problem? Because a large correction usually compensates for a mechanical misalignment or a bad fiducial. Fix the alignment or the target first, then apply the small remaining correction.

Should offset be measured on every board of a panel? Measure at more than one location, including the panel edges. Panel growth can leave the middle boards centred while the outer boards are offset.