Adom Custom Electronic Load

@@ -61,3 +61,11 @@ The actual portable C controller and tach decoder run with synthetic inputs. At <video width="100%" playsinline controls preload="metadata" poster="docs/clips/fan-safety-poster.jpg" src="docs/clips/fan-safety-replay.mp4"></video>  The recording preserves the simulation label. Recorder timing repair failed; use the displayed simulation clock for event timing, not encoded-video duration. [Replay source and trace](../demos/fan-safety/README.md).++## Routed prototype: native 3D review++Fresh top, oblique and bottom views of the saved routed board, including its prototype label and copper. This is an independent review take after routing, not a reconstruction of the missed route-growth footage. An 8.2-second excerpt with no speed change; the raw recorder timing warning is retained. No cooling assembly or hardware qualification is shown.++<video width="100%" playsinline controls preload="metadata" poster="docs/clips/routed-3d-review-poster.jpg" src="docs/clips/routed-3d-review.mp4"></video>++![Reviewed native views](docs/clips/routed-3d-review-sheet.jpg)

⋯ 1 unchanged line ⋯

⋯ 1 unchanged line ⋯

⋯ 1 unchanged line ⋯

@@ -4,9 +4,9 @@ The accepted plan has landed in native KiCad: 230 route operations, 1,674 track  This is a routing checkpoint, not a fabrication release. The 5 W bring-up ceiling remains. USB pair/reference review, power-path/current screening, cooling geometry, model qualification, final silk and hardware calibration are still required. No hardware has been assembled. The 25 W target is unvalidated. -The installed Fields 0.6.2 solver incorrectly assumed four copper layers. Its run was stopped and its partial results are not accepted. A two-layer correction passes 18 tests and is submitted as Fields PR 4; it has not replaced the shared installation. Its actual-board 0.2 mm solve is rejected for GND DC nonconvergence; see [field-screen review](fields-review.md). Final current CAD views are also pending: native save/DRC evidence is available, but Windows currently returns stale or blank canvases.+The installed Fields 0.6.2 solver incorrectly assumed four copper layers. Its run was stopped and its partial results are not accepted. A two-layer correction passes 18 tests and is submitted as Fields PR 4; it has not replaced the shared installation. Its actual-board 0.2 mm solve is rejected for GND DC nonconvergence; see [field-screen review](fields-review.md). Current PCB and 3D views now repaint correctly and have been visually verified, including the saved prototype title and routed copper. -There are 26 raw video takes and nine reviewed excerpts. Two new recordings ran during actual routing, but pixel checks found stale canvases and they are excluded from the action edit; the fan-fault excerpt is clearly labeled software simulation. The final narrated 4–5 minute film is not complete.+There are 27 raw video takes and ten reviewed excerpts. A fresh native routed-board 3D review now shows verified changing top/oblique/bottom views. Two new recordings ran during actual routing, but pixel checks found stale canvases and they are excluded from the action edit; the fan-fault excerpt is clearly labeled software simulation. The final narrated 4–5 minute film is not complete.  Evidence: [saved-board DRC](../design/routed-live-drc.json), [native net audit](../design/accepted-native-net-audit.json), [ground correction](../design/ground-routing-review.md), [recording inventory](../videos/inventory.json). 

@@ -1,6 +1,6 @@ {-  "rawTakeCount": 26,-  "reviewedExcerptCount": 9,+  "rawTakeCount": 27,+  "reviewedExcerptCount": 10,   "countingRule": "Unique raw window recording filenames in this project; excludes derivative action/speed exports and reference projects. Raw count includes rejected takes.",   "takes": [     {@@ -158,6 +158,12 @@       "durationSeconds": 475.7,       "topic": "Native route landing with stale captured canvas; not usable construction footage",       "review": "Excluded from action edit: sampled board pixels are unchanged while status changes; stale canvas."+    },+    {+      "file": "videos/raw/window-4655002-20261008-170541.mp4",+      "durationSeconds": 10.9,+      "topic": "Fresh routed prototype native 3D top, oblique and bottom review",+      "review": "contact sheet reviewed; changing views verified; 8.2-second excerpt decoded successfully; recorder timing warning retained"     }   ] }

@@ -164,6 +164,20 @@       "status": "Excluded from action edit: sampled board pixels are unchanged while status changes; stale canvas.",       "note": "Planning and bridge waits must be removed. Part 2 shows trace additions more clearly; old UNROUTED title remained during routing and was corrected after saving.",       "review": "Excluded from action edit: sampled board pixels are unchanged while status changes; stale canvas."+    },+    {+      "id": "routed-3d-review",+      "file": "raw/window-4655002-20261008-170541.mp4",+      "topic": "Native routed-board top, oblique and bottom inspection",+      "status": "reviewed excerpt; full decode passed",+      "duration_s": 10.9,+      "excerpt": "../docs/clips/routed-3d-review.mp4",+      "sourceIntervalSeconds": [+        1.6,+        9.8+      ],+      "speedChange": 1,+      "note": "Independent review take after routing, not construction footage. Recorder timeline repair warning retained. No fan/heatsink assembly shown."     }   ],   "failed_captures": [