Manufacturers should characterize I-Tera MT40 dimensional compensation under their own material structures and process conditions. The manufacturer's guide explains that net inner-layer dimensional change after etching, surface treatment, and lamination generally appears as shrinkage. Warp/fill directions, remaining-copper distribution, and lamination methods affect results. Compensation is not a constant determined by material grade alone.
Lock measurement baselines and material orientation first
Select datums covering the panel center and perimeter, retaining point IDs, design coordinates, measurement equipment, environment, and process state. Core and PP warp directions must remain traceable through cutting and layup records; do not infer direction from the outline's long edge after measurement. Comparing the same points after imaging, etching, and lamination identifies which process causes changes.
The guide updated on 2025-12-8 recommends characterizing movement of unbaked material and states that each factory should determine compensation using its lamination cycles, border designs, and material directions. Reference tables specify copper weights and hydraulic-press conditions; one row cannot directly apply to orders with different copper distributions. Heavy-copper structures with extensive copper removal particularly require review.
How to calculate compensation coefficients for contraction
In an engineering example, two warp-direction datums have a design spacing of 400.000 mm and measure 399.680 mm after processing. Relative change is (399.680－400.000)÷400.000＝－0.0008, or －0.08%. Under a simple uniform-scaling model, the next artwork enlargement factor is approximately 400.000÷399.680≈1.000801, or about 0.0801% enlargement.
Measure the fill direction independently; do not automatically reuse the same factor. Before calculation, confirm calibration and temperature effects and exclude coordinate-alignment errors. If corner deviations point in different directions or the center is correct while edges shift, one scale factor may be insufficient. Check rotation, translation, trapezoidal distortion, and local deformation first. Do not keep increasing global compensation to conceal uneven local lamination.
| Measurement item | Sample Record | Corresponding Judgment |
|---|---|---|
| Warp-direction distance | 400.000→399.680 mm | Estimate warp-direction scaling, retaining process stage and direction |
| Fill-direction distance | Measure Same Batch Reference Points Separately | Establish independent scaling and variation |
| Diagonal and corner residuals | Re-Compare After Scaling Correction | Identify rotation, distortion, and local offset |
| Interlayer Alignment | associate with drilling reference and targets | Determine whether improved artwork dimensions improve connections |
Lamination parameters must follow the structure
The guide distinguishes standard and sequential lamination and notes different cure-time needs for thick boards. Process records should distinguish press setpoints from actual product temperatures and retain heating, vacuum, pressure, dwell, and cooling stages. The peak shown on the press screen is not the material's complete thermal history.
Material-combination changes, asymmetric copper patterns, PP ply counts, and border designs can change resin flow, dielectric thickness, and dimensions. Fix one combination first in trials, confirm fill and interlayer bonding, then evaluate compensation. Changing copper weight, PP, and scaling simultaneously produces an unexplained “best value” that is difficult to reproduce on reorders.
- Assign route IDs to each substructure and final lamination, listing materials, directions, copper patterns, and target dielectric thicknesses.
- Record product temperatures at representative locations and batch information; check applicability of manufacturer guidance.
- After lamination, inspect thickness, fill, edges, and registration. Dimensional acceptance alone does not establish interlayer quality.
- Feed actual results into compensation files for the corresponding structure, retaining change rationale and applicability.
Do not simply reuse ordinary FR-4 drilling and edge-trimming windows
The guide gives more conservative initial cutting speeds and feeds than conventional FR-4 and requires adjustment for total thickness, layer count, copper weight, glass construction, and tool condition. Identical hole diameters do not establish identical tool life at different stack heights. Drilling records should include tool diameter, spindle speed, feed per revolution, stacked-board count, entry/backup materials, and cumulative hole count.
Evaluate hole-wall roughness, resin residue, drill wander, and exit damage alongside tools and material structures. Dimensional issues may also arise from inconsistent drilling and artwork datums. The guide recommends routing lamination resin flash to reduce edge damage. Validate specific tools and routes with samples so dimensional improvements do not create new edge cracks.
Close abnormalities against the same measurement points and via structures
| Anomaly | Troubleshoot first | Closure Evidence |
|---|---|---|
| Overall Uniform Reduction | Directions, structures, process shrinkage, and scaling revisions | Re-measurement of grouped points and residual after compensation |
| Local Hole Alignment Error | Drilling datums, local deformation, and interlayer movement | Target measurements, representative cross-sections, and connection checks |
| Dielectric Thickness and Impedance Shift Simultaneously | resin flow, residual copper, and PP combination | Cross-section dielectric thickness and test coupon at same location |
| Panel Edge Anomaly | Borders, edge trimming, and pressure/temperature uniformity | Location map, edge images, and re-inspection |
The final output is measurement data, calculation tables, compensation files, material lots, and applicable structures, not a context-free percentage. New batches can use representative measurements to check historical models; recharacterize changes beyond validated scope. Design teams should receive stable manufacturing stackups and permitted width-adjustment ranges, avoiding confusion between manufacturing scaling and arbitrary circuit resizing.
How to manage compensation for multiple laminations
In sequential lamination, inner substructures and final boards experience different thermal histories. Record dimensions and datums before and after each cycle separately, identifying layers reheated later. Final-outline dimensions alone cannot show whether via-layer pairs shifted relative to each other during an earlier cycle or whether artwork, drilling datums, or layup directions need changes.
Compensation files need names, revisions, generation bases, approval dates, and covered structures, with machine programs referencing identical IDs. Retain old and new factors rather than overwriting same-name shop-floor files. Initial-batch retesting should assess whole-panel residuals, not just the two points closest to target. Investigate instability before averaging widely scattered data, which can hide local abnormalities.
Customer reports may summarize critical dimensions, registration, and impedance, while internal raw coordinates should remain linked to batches. External acceptance tolerances, internal process-control limits, and manufacturer reference starting points are different objects. Label them separately to prevent process-adjustment references from becoming purported approved product tolerances.
How to avoid hiding compensation problems with averages
Retain raw coordinates for every panel and measurement point of the same structure. Check consistency before calculating average factors. Different corner-deviation directions or a clearly shifted batch may not reflect uniform shrinkage. Averaging all data directly can produce a compensation value applicable to no actual board.
After model compensation, check point residuals and spatial distribution, recording what the model explains and what remains. For poor measurement repeatability, first check clamping, calibration, and temperature. For large material/process variation, improve stability first. Compensation corrects identifiable systematic changes; it cannot replace a stable process.
Approved files can define applicable cores, copper distributions, and directions, plus changes requiring remeasurement. Do not automatically reuse old values if sample and production routes differ, lamination equipment changes, or layup combinations change significantly. Focus reorder checks on these triggers to reduce unnecessary repeated measurement.
Frequently Asked Questions
Can pre-baking eliminate all dimensional drift?
The manufacturer does not present stress-relief baking as a universal remedy and still emphasizes measured material movement. Baking also involves moisture and surface-treatment conditions and requires process-specific review.
Can original compensation be reused for the same board thickness?
Reuse is supported only when material structures, copper distributions, directions, and lamination routes fall within validated ranges. Refer to Manufacturing Combination with Manufacturing Document Checklist Retain Complete Records.
Source material and applicable scope
- Isola Group, “I-Tera MT40 Processing Guide”; version: Last Updated December 08, 2025; checked 2026-10-17; relevant sections: Part 2 Innerlayer Preparation, Part 3 Lamination, Part 4 Drill. Supports factory-specific characterization of directional dimensional changes and structure-dependent lamination/drilling adjustments. Numerical examples are this site's demonstrations; reference compensation tables are not factory guarantees.
