Printed Resistor Geometry: Separating Wet Spread from Fired Film Change

Define geometry at four process states

Record artwork or screen aperture, wet deposit after a controlled delay, dried film before firing and fired film after conditioning. Use width, length, corner and overlap features that answer the design question. Wet paste can slump with time and orientation, so state imaging delay and handling. Fired optical contrast may reveal a glassy edge different from the electrically active boundary. Inspection should use a controlled threshold or reviewed manual rule rather than whichever edge is easiest to see.

Calculate stage-to-stage dimensional bias

Normalize positions to substrate fiducials to remove rigid registration, then compare widths or edge coordinates between adjacent stages. Report left and right edges separately before collapsing them into width.

Distinguish spread from later contraction in one example

For illustration, a 0.80 mm aperture produces a 0.86 mm wet deposit, 0.84 mm dry feature and 0.82 mm fired feature. Wet bias is +0.06 mm; dry-to-fired change is −0.02 mm; final bias to the aperture is +0.02 mm. Calling the result two-percent shrinkage would hide the earlier spread and use an ambiguous denominator. These values demonstrate decomposition only and are not process capability or paste data.

Use fiducials, repeated sections and edge profiles

Measure across straight bodies, ends, corners and conductor overlaps because flow and firing effects differ. Include orientation relative to print direction and screen travel. Use common fiducials to align images and measure substrate dimensions before and after firing. Cross-section selected specimens where thickness or edge taper matters, recognizing that section preparation is destructive. Calibrate imaging scale and estimate repeatability by remeasuring images. A single field of view does not represent panel position or screen wear.

Vary print and firing factors without confounding them

Screen mesh, emulsion, snap-off, squeegee, paste conditioning and print speed influence wet transfer. Drying and firing profiles influence consolidation and edge appearance. To identify an interaction, hold the other stage constant or use a planned matrix. A new screen often changes aperture and transfer simultaneously; measure both. Paste lot or idle time should be traceable. Do not attribute a fired width shift to kiln conditions when wet images already contain the same shift.

Relate geometry to resistance through a separate validated model

Length and width influence nominal squares, but edge taper, thickness, conductor overlap and local current field affect the electrical result. Use measured fired geometry in an applicable resistor model and compare four-terminal or defined terminal resistance. A geometry match with changed resistivity is possible, as is a geometry change with little net resistance change. Preserve these as separate outcomes. Laser trimming, if used later, creates another state that must not be folded into as-fired geometry.

Prevent false corrections from optical and substrate artifacts

Illumination, focus, glaze, porosity and threshold can shift apparent edges. Substrate expansion or camera calibration can make every feature seem larger. Registration drift moves both edges without changing width. Inspect fiducials, scale references and raw profiles before changing artwork. Overcorrecting aperture for a transient paste or screen condition can create a durable future bias. Any process adjustment should be validated across relevant panel locations and repeated builds.

Control final geometry and stage evidence without inventing a universal bias

The drawing should define the fired functional boundary, datums, conductor overlap and measurement rule. Process records should link artwork, screen, paste, print settings, dry and firing profile. ChipSimple can establish internal controls for an agreed construction after applicable evidence; no fixed spread factor applies to every paste, mesh or shape. Reopen characterization after screen, paste, profile, substrate, feature scale or inspection-method changes.

Measure the printed body as a fired process result

Artwork dimensions describe the opening presented to a new screen; they do not directly describe the fired resistor. Paste transfer depends on mesh, emulsion, squeegee setup, snap-off, substrate surface and paste condition. Wet deposits can spread after separation, while drying and firing change both lateral edge position and film thickness. The review should therefore distinguish artwork length, wet print extent, dried extent and fired effective dimensions. Termination overlap also shifts the electrical boundary. A microscope image of the green body alone cannot establish the number of effective squares unless conductor interfaces and edge-definition rules are visible.

Build a coupon map that crosses print direction with panel position. Measure width at several stations, including near conductor overlaps and any neck or corner, rather than reporting one centre value. Record fired thickness with a method suited to the film and substrate; surface texture can make a single trace misleading. Pair geometry with resistance before trimming. If the same nominal artwork produces a systematic leading-edge or trailing-edge difference, investigate separation mechanics and print direction before changing every resistor target. Statistical summaries should retain lot, screen, firing run and measurement orientation so process effects are not pooled into an anonymous distribution.

When a correction is needed, decide which controlled object carries it. Artwork compensation can address stable lateral bias, but it should not mask unstable transfer or screen wear. Paste or firing changes can alter sheet resistance and thickness as well as geometry, so a dimensional correction alone may move electrical results unexpectedly. Request the resistor material family, print route, termination layout, available coupon space and initial-versus-final resistance window. Release a geometry only after representative fired results demonstrate that its nominal target and tolerance robustness coexist under the reviewed route.