12 V to 5 V / 1 A Buck Molecule (TPS54202)

@@ -23,6 +23,10 @@ Watch the board being built in KiCad 10, step by step. Each player below runs at | 15 | [First silkscreen pass (first run)](#clip-15) | | 16 | [First 3D walkthrough (first run)](#clip-16) | | 17 | [Probing sends the board back: TP5](#clip-17) |+| 18 | [Probing sends the board back: a VIN test point (TP6)](#clip-18) |+| 19 | [Collision check: a ground contact the pincer can reach, two test points](#clip-19) |+| 20 | [Labels for every probe point](#clip-20) |+| 21 | [Probe pads out from under the silkscreen](#clip-21) |  ## 1. Placement @@ -196,3 +200,35 @@ The probing plan found that TP3, the ripple test point, had no ground pad within  <video controls preload="metadata" width="1080" poster="docs/probing/tp5-loop-poster.jpg" src="docs/probing/tp5-loop.mp4"></video> +## 18. Probing sends the board back: a VIN test point (TP6)++<a id="clip-18"></a>++Working out what this board must have measured showed that 12 V had no test point on the board: the only place to land on VIN was the big input capacitor's pad, right beside its 5.8 mm can. The run went back to the layout step and TP6 was added in KiCad 10. The first spot was refused for crowding two parts, the straight trace for crossing EN, and the next route for passing too close to a ground strap; the third route was clean. Then the pours refill and VIN reaches TP6. Placement, routing and pours, in that order.++<video controls preload="metadata" width="1080" poster="docs/probing/vin-tp6-loop-poster.jpg" src="docs/probing/vin-tp6-loop.mp4"></video>++## 19. Collision check: a ground contact the pincer can reach, two test points++<a id="clip-19"></a>++The probing simulation moved the workcell's head through the whole test plan against the board's real 3D parts and the wires to the control panel. Three things would not have worked on the real cell. The pincer that plugs the probe's ground pin in would have hit the wires on the contacts either side of J5. The inductor's output pad sits under its body. And the input capacitor's end cap left too little of its pad for the probe tip. Each fix was checked against the same geometry before it touched the board: J7, a ground contact with room for the pincer's jaws, and TP7 and TP8, test points beside the input capacitor and the inductor's output. They land one at a time, the pours reach them, and the board is saved.++<video controls preload="metadata" width="1080" poster="docs/probing/collision-redesign-poster.jpg" src="docs/probing/collision-redesign.mp4"></video>++## 20. Labels for every probe point++<a id="clip-20"></a>++Every test point and contact the workcell uses gets its reference, and its net where there is room, at the smallest text this fab prints that is still readable (0.32 mm high, 0.08 mm stroke). Each label is placed beside its own pad and checked against the pads, the other text and the silkscreen lines before KiCad places it. Shown at twice real speed.++<video controls preload="metadata" width="1080" poster="docs/probing/probe-point-labels-poster.jpg" src="docs/probing/probe-point-labels.mp4"></video>++## 21. Probe pads out from under the silkscreen++<a id="clip-21"></a>++The fab clips silkscreen wherever the solder mask is open, so text over a test point prints broken and the pad carries ink at its edge. Three of the new pads had landed under existing labels, so they move to clear spots, keep their names, and take their labels with them. The ground contact moves to where the pincer still has room. The input capacitor also gets its updated 3D model. TP5, the ground pad from clip 17, is no longer needed (the probe grounds through J7) and comes off, which leaves room for TP3's label beside its pad. Shown at twice real speed.++<video controls preload="metadata" width="1080" poster="docs/probing/silk-clear-pads-poster.jpg" src="docs/probing/silk-clear-pads.mp4"></video>+

⋯ 1 unchanged line ⋯

⋯ 1 unchanged line ⋯

⋯ 1 unchanged line ⋯

⋯ 1 unchanged line ⋯

⋯ 1 unchanged line ⋯

⋯ 1 unchanged line ⋯

⋯ 1 unchanged line ⋯

⋯ 1 unchanged line ⋯