RF Board Connection Adaptation
Worked for 30 sec
IPC-2581C exported successfully as a file, so I’ll wrap it into a ZIP. DXF export failed once; I’m retrying DXF before packaging.
Thought for 7 sec
DXF exported successfully as a ZIP. I’m wrapping the IPC-2581C file into its own ZIP now.
Thought for 20 sec
The zip wrapper tool didn’t receive the IPC attachment automatically, so I’m doing one direct download-and-zip attempt from the signed IPC-2581C URL.
Thought for 35 sec
DXF ZIP is ready:
IPC-2581C exported successfully, but Flux produced it as a single
.ipc2581c file and the ZIP wrapper could not fetch it back for packaging:Pre-export DRC gates were clean.
Package and verify exports
Export ODB++ too
Export STEP model
Check package contents
Create fab email
Refine this doc
Ask about this doc
Final DRC / SI Cleanup Status
Date: 2026-07-19
Actions performed
- Removed the bad overlap-producing J2 upper escape route and its dangling remnants.
- Reran autoroute cleanup passes after removing the bad route fragments.
- Repaired temporary 3V3/WSI dangling segments caused by cleanup attempts and restored routing to the current post-autoroute baseline.
- Ran a targeted full DRC/SI review set covering airwires, overlaps, dangling traces, invalid layers, footprint/body/board checks, pad access, stackup validation, and selected impedance/differential-pair/SI checks.
- Visually reviewed the PCIe/refclock route corridor between J1, AC-coupling caps, and J2.
Current review result
The latest review reports no open overlap, dangling-trace, protected-intrusion, invalid-layer, stackup, component-body, component-outside-board, pad-access, or selected SI/impedance warnings/errors.
Open DRC items are limited to three J2/GND airwires:
Table
| Check | Endpoints reported |
|---|---|
401437a5 | GND stitch via via.stitch.36 to J2 GND_15 |
GND_65 | J2 GND_65 to J2 GND_67 |
GND_76 | J2 GND_76 to J2 GND_74 |
Notes / risk
- These three remaining items are connector-ground stitching/local ground continuity items on the SK2X40 J2 footprint.
- Direct pad-to-pad routing and autoroute could not resolve them cleanly.
- Manual multi-segment jumpers can clear the airwire checks, but they caused overlapping-copper DRC at segment joints or nearby existing copper, so they were removed to avoid hidden fabrication/short risk.
- J2 GND pads already have direct/all-layer connection rules on the stubborn pads, but the checker still reports the above airwires.
Recommended next action
Do not fabricate until these three J2 GND DRC items are resolved or formally accepted as same-net false positives after manual CAD review. Preferred next repair path is to edit the J2 footprint/copper geometry directly with a single continuous same-net copper shape per jumper, or update the connector footprint/pad access so the ground pads connect cleanly without trace-segment overlap reports.
Actions performed
Current review result
Notes / risk
Recommended next action
