Tracing Panel Scaling in PCB Direct Imaging

Panel scaling in direct imaging is a controlled manufacturing transformation. When registration varies, the useful evidence includes target locations, measured panel behavior and the selected transformation. A screenshot showing a visually improved overlay is not enough to define what changed.

Identify the available transformation

KLA’s Nuvogo 80F brochure describes multiple scaling modes and panel traceability functions. These equipment capabilities illustrate that scaling can be more specific than a single global percentage; the actual process must identify the mode and configuration used.

Record the CAM source revision, imaging job, target definitions and orientation. Keep the measurement coordinates tied to the physical panel so a later comparison uses the same reference system.

Separate the correction from the product requirement

Check which parts of the product are constrained by the same transformation. Improving agreement near selected targets does not automatically establish every dimension or interlayer relationship elsewhere on the panel.

Use a spatial comparison

Map observations across the panel and retain signs and directions of deviations. An unsigned average loses information about whether features shift together or diverge in different regions.

Review construction and processing history when patterns change between lots. A scaling value copied from a previous job should not conceal a different material or intermediate-state condition.

Release the correction with evidence

1. Define the problem and affected feature relationships.
2. Retain baseline measurements and transformation settings.
3. Evaluate the proposed correction on representative panels.
4. Inspect critical locations using the agreed method.
5. Document the approved scope and review triggers.

Keep customer design data distinct from imaging corrections. If the intended final geometry changes, update the design release explicitly rather than hiding that change inside a manufacturing transform.

For an unexplained registration issue, preserve rejected or marginal panel data where available. A study containing only accepted panels can omit the information needed to understand the process boundary.

The final record should let another qualified reviewer reconstruct the relationship between source data, observed panel and applied correction. That is the practical difference between a repeatable imaging process and a setting chosen from memory.

For example, label a measurement taken from the opposite panel side with its actual viewing convention. Reversing a coordinate or sign during data transfer can make a real positional trend appear inconsistent, leading the team to change imaging settings before resolving the reference error.

See position-reference definitions for coordinate control and CAM compensation traceability for a related manufacturing transformation.