On this page
An uncovered window needs more than a nominal gap between programmed coating passes. Placement error, edge spread and the rule used to detect a thin residue can all reduce the region that remains usable. Work outward from the required clear feature and assign a separate inward allowance to each adjacent coating boundary. The result is a geometric requirement to validate on the actual coated assembly.
Key design decisions
- Define the fully uncovered functional region before drawing the coating path.
- Allocate inward movement separately for each edge and deposition direction.
- Inspect the final material boundary, including residue outside the visually dense film.
Define what must remain accessible inside the window
A probe land, a connector contact and a bonding region can require different meanings of uncovered. One needs electrical contact over a specified footprint; another may prohibit any polymer on a sliding surface. Identify the functional region and the stage at which it must be available. A gap that is open immediately after dispensing may receive material from later flow or another coating operation. Residue in that region creates a contamination risk at the contact even when the surrounding protective film appears continuous.
Mark the contact or tool envelope on the circuit drawing with its datum relationship. Include the area the tool actually touches and any surrounding region whose contamination affects the operation. Keep this requirement distinct from the path used to apply coating. The applicator centerline is a manufacturing input, while the final empty region is the result being controlled. Stating the result first makes it possible to compare different application routes without changing the electrical or assembly requirement whenever a path is adjusted.
Distinguish the dense film edge from the last unwanted material
A coating pass can have a region of full coverage, a less uniform boundary and isolated material beyond that boundary. An edge measured at the dense film contour may therefore overstate the fully uncovered width. Establish which contour matters to the contact function and how it will be detected. Transparent residue may need an appropriate compatible inspection method; absence of an obvious colored edge is insufficient evidence.
Retain the measurement criterion when comparing dispensing conditions. A setting that narrows the dense stripe while increasing satellite droplets can appear to improve the window under the wrong metric. Record isolated incursions separately from continuous edge movement because they may require different corrective actions. An allowed thin residual film, if any, needs an explicit functional basis. Do not convert an instrument detection limit into permission for coating on a surface that the drawing requires to be completely exposed.
Build a separate inward allowance at each coating boundary
For each side, identify the largest credible movement toward the protected opening. It can include position error relative to the circuit datum, fluid landing offset and additional inward spread before the evaluated final state. Use distinct terms only when their boundaries are distinct. If a measured final-edge distribution already includes landing and spreading effects, adding those effects again would count the same motion twice.
The left and right sides need not have equal allowances. Different travel directions, neighboring steps or application passes can change their behavior. Retain the direction and sequence behind each value. Include an explicit measurement or decision margin where the acceptance method requires one, with its purpose identified. The budget should distinguish measured process bounds from design reserves and unresolved inputs. A single symmetric tolerance copied from an equipment brochure does not establish the final material boundary beside this particular ceramic feature. A common registration shift can relocate the whole gap without narrowing it, so its position relative to the required contact must also be checked.
Calculate the nominal gap required to preserve the clear width
Let the nominal gap lie between the intended coating boundaries before their inward allowances are applied. Subtract the left and right allowances to obtain a conservative remaining width. If the design starts with a required uncovered width, add both allowances to that width to determine the necessary nominal gap. A negative remaining width means that the assumed bounds permit the coating regions to meet or overlap.
In an illustrative design, require a clear width of 1.20 millimeters. On the left, assume 0.08 millimeter of placement allowance, 0.10 of inward edge spread and 0.02 of decision margin, totaling 0.20. On the right, assumed terms of 0.06, 0.08 and 0.02 total 0.16. The nominal gap must therefore be at least 1.56 millimeters. Reserve 0.20 to the left and 0.16 to the right of the required region; centering this gap would provide only 0.18 on each side and miss the left-side allowance. A 1.50-millimeter nominal gap would leave only 1.14 millimeters under these bounds, short of the illustrative requirement by 0.06 millimeter. No value here is a dispensing capability or recommended production allowance.
Gclear,min = Gnom − aleft − aright; Gnom,required = Wrequired + aleft + aright
- Gnom: intended uncoated gap between nominal final-edge targets, in mm
- aleft and aright: separately justified maximum inward allowances, in mm
- Wrequired: required clear functional width, in mm
- Gclear,min: conservative remaining width under the stated bounds, in mm
Locally opposing edges evaluated in one direction with positive inward allowances and consistent datums. Each physical error is counted once. Corners, isolated droplets and nonparallel boundaries require separate geometric checks. Width alone is insufficient for a fixed target: verify each edge against its required target coordinate.
Check pass direction, starts, stops and corners independently
A long straight pass can be easier to characterize than its start, stop or turn. Material released from a moving applicator need not land directly beneath its release position, and a changed direction can change the sign of its offset relative to the opening. Retain the actual path and motion state for the boundary measurements. Applying one correction to every direction can protect one side while moving another toward the contact.
At a corner, two individually acceptable straight boundaries can still leave a rounded or locally filled region inside the required tool envelope. Overlay the complete measured contour on the required region instead of checking width only at the middle. Nearby component feet, trenches or surface steps can also redirect liquid. A flat witness is useful for separating delivery behavior, but the corner beside an assembled feature needs a representative geometry check before the calculated allowance can support the finished drawing.
Keep allowance evidence tied to the mechanism it represents
An allowance becomes reviewable when its derivation and measurement state travel with the number. State whether it comes from calibrated position uncertainty, observed final-edge excursions or a deliberate design reserve. A maximum observed excursion from a limited sample does not automatically bound every future part. The validation plan must establish the population and conditions that the selected engineering bound is intended to cover.
The table helps prevent two common errors: using equipment position repeatability as if it measured fluid spread, and treating an optical boundary rule as if it were a process variable. Both can reduce the practical clear region, but they enter the review through different evidence. If one contribution dominates the required gap, examine whether a changed path, local mask or layout can address it. Such a change still needs confirmation of coverage on the opposite side of the boundary.
| Budget item | What it represents | Record required |
|---|---|---|
| Datum-relative positioning | Location of the intended boundary on the actual circuit | Fiducial definition and measured placement error |
| Moving-fluid offset | Landing displacement associated with delivery motion | Direction, height and start/stop comparison |
| Post-delivery spread | Material motion before the accepted final state | Registered wet and cured boundary observations |
| Detection or decision margin | Unresolved boundary measurement or acceptance reserve | Inspection method and margin rationale |
| Local geometry exception | Corner, step or isolated intrusion | Complete contour versus functional envelope |
Use the shape of a window failure to choose the correction
A parallel shift of an otherwise unchanged coating edge suggests a different investigation from a widening diffuse boundary. A consistent incursion at pass ends directs attention to start and stop behavior. Isolated droplets in the clear area can escape a width-only inspection entirely. Preserve the spatial pattern and the application sequence before wiping or stripping the failed region.
Correlate the observation with actual contact behavior where the application permits a controlled test. An intermittent probe reading can arise from polymer residue, contact geometry or the probe itself, so an electrical symptom alone does not identify coating spread. Keep an uncovered control and verify the contact setup. If an intrusion appears only after cure, compare the earlier image using the same coordinate registration and boundary criterion. A changed lighting threshold must be excluded before concluding that the material physically moved during the thermal step.
Validate the final window and the protection immediately outside it
Run the complete proposed coating sequence on representative features, including the least favorable corners and nearby steps. Measure the final clear contour after the agreed cure and conditioning state. Demonstrate that the required contact or access region fits within it under the chosen acceptance rule. Use independent coated assemblies or preparations appropriate to the process question; repeatedly measuring one attractive window does not establish deposition repeatability.
Also examine the region that must remain protected. Moving a coating path outward to rescue a contact can expose a resistor edge or create a gap between passes. Qualification must preserve both requirements simultaneously. The handoff should contain the required empty region, nominal coating boundary, directional allowance budget, inspection method and identified exceptions. Keep future changes in coating formulation, application path, feature height or cure route connected to the terms they affect. This produces a usable window allowance without claiming that a nominal machine path guarantees a clean contact.
Provide the uncovered region and coating boundary data
Send the contact function and directional edge evidence needed to size the coating exclusion.
- Drawing with the required clear contact or tool envelope, circuit datums and neighboring protected features.
- Named polymer, deposition method, complete pass path and the wet-to-cured process sequence.
- Directional placement and final-edge measurements with starts, stops, corners, steps and isolated incursions identified.
- Boundary detection method, input-bound definitions, proposed decision margin and representative contact or access validation.
The drawing-upload form loads as you reach this section.