吃瓜tv

Establish which files and environment are being checked

Record the EAGLE version and preserve the paired schematic and board before making repairs. Confirm that they describe the same intended design state. A board can pass geometric rules while implementing a different revision from the schematic, so source pairing needs its own review.

Identify the requested construction and receiving requirements. A two-layer prototype and a multilayer controlled-impedance design do not share every relevant constraint. Enter justified rules for the project rather than relying on an arbitrary downloaded rule file whose scope or date is unknown.

If a manufacturer provides a rule file, keep its source and applicability visible. It can assist layout checks but may not encode every special construction or assembly requirement. Confirm unresolved limits with the receiving team rather than interpreting a file as a blanket capability commitment.

Review the rule configuration before the error list

Check that the rules reflect the intended layer stack, dimensions and feature requirements. Clearance, size and distance checks only have meaning relative to the configured values and scopes. A mistakenly generous rule can produce many false concerns; an overly permissive rule can hide a real issue.

Keep units explicit when transferring limits from drawings or supplier notes. The mm-to-mil calculator can help convert a number, but it cannot decide which requirement should govern. Rounding a minimum downward can change the constraint, even when the difference looks small on screen.

Document local exceptions with a reason and an owner. A deliberate feature may need an agreed interpretation, but that is different from globally relaxing a rule to remove one warning. Retain enough context for a later engineer to understand why an exception remains acceptable after a component or stack change.

EAGLE DRC Before Gerber Export review

Classify findings by the underlying problem

Separate connectivity concerns from geometric spacing, outline proximity, hole definition and documentation problems. A list ordered only by error count can encourage easy cosmetic fixes while leaving a consequential issue unresolved. Start with findings that challenge the intended circuit or physical construction.

For a spacing violation, identify the actual objects, nets, layer and required limit. For a hole concern, identify size, plating and span rather than treating every circular feature identically. For an edge issue, establish the finished outline and relevant physical tolerance. A screenshot without these details often leaves the reviewer guessing.

Some findings require a design change, while others require corrected data or an explicit exception. Record the action and rerun the relevant check after making it. Do not mark a finding resolved solely because the object was hidden or the message was dismissed.

Recheck generated copper and source consistency

Copper pours and other generated objects can change after edits. Regenerate using the appropriate software workflow, inspect the resulting shapes and rerun the relevant checks. A layout that looked clear before regeneration may develop narrow regions or altered return paths afterward.

Review electrical return and power paths as engineering requirements, not just clearance geometry. A geometrically legal route can still be unsuitable for the circuit. DRC can support that review when the relevant constraints are modeled, but it cannot infer every unstated signal or power requirement from a net name.

If the schematic changes, reconcile the board deliberately and review the resulting modifications. A renamed net or substituted part can affect more than the visible connection lines. Preserve the change record so the export can be tied to the correct source revision rather than an uncertain intermediate state.

Verify the exported package as a separate stage

After the native design review, generate a coherent manufacturing archive and inspect it independently. Check dimensions, layer function, drills, mask openings and the intended profile. Passing native DRC does not ensure that the CAM job selected the correct layers or included every required file.

A worked example is a board with correct hole clearances in the editor but a drill file generated with a different origin. The source DRC can be clean while the exported overlay is wrong. This is an output configuration problem; changing the layout geometry to compensate would obscure the cause.

Another example is a current board exported alongside an older inner-layer file left in the destination directory. The number of files may look right. A clean export destination, manifest and independent content comparison provide evidence that DRC alone cannot supply.

Turn the checklist into a useful handoff

The worksheet asks you to record the legacy environment, source consistency, rule basis, issue disposition and output agreement. Its states are self-reported. A zero-open count means you selected evidence states; it does not mean the website verified your design or that a manufacturer approved the construction.

For an unresolved item, write the affected feature, intended requirement, observed behavior and requested decision. Include a marked view and the relevant revision through the agreed project channel. Keep sensitive customer content out of public examples or tool notes.

Use the PCB DFM page for the broader manufacturing review and the keepout guide for the distinction between geometry restrictions and physical features. The objective is a traceable path from requirements through native checks to inspected files, with unresolved engineering questions clearly identified before quotation or production.

Keep an archive decision log

Record the original environment, source identity and the evidence supporting each interpretation. Note whether a result came from an intact native project, a migration, a fabrication archive or a drawing reconstruction. Those starting points preserve different information and should not receive the same approval label.

For each unresolved item, identify the missing evidence and the consequence of proceeding without it. A successful import can be an intermediate milestone while plane behavior, library mapping or output settings remain uncertain. Retain the original and give any repaired or migrated derivative a distinct identity. When a later reviewer opens the project, the record should explain what was changed deliberately, what was merely translated and what still needs a decision.