Top Silkscreen Layer

d6409383-ab7d-4768-bdde-3a6311d6f610 is this on the top or the bottom layer?

Thought

flux

d6409383-ab7d-4768-bdde-3a6311d6f610 (“Credit - Name” / “Brooks Solveig”) is on the Top Silkscreen layer.

Files

Schematic

Layout

3D View

Buck Layout Independent Review

Reviewed: 2026-09-10 after buck-converter layout rework. Scope: U3 (AP63203WU-7), L1, C3, C12, C19, and the aggregate 3V3 trunk.

Result summary

Completed — 3V3 budget: schematic net properties are 0.365 A estimated and 0.5 A maximum. U3 is rated 2 A. L1 is 3.3 µH, 9.5 A saturation / 8.5 A rated; 3.3 µH is inside the datasheet’s approximate 2.2–10 µH usable range, although the fixed-3.3 V recommendation table lists 3.9 µH. At 5 V in, 3.3 V out, 1.1 MHz and 3.3 µH, calculated ripple is about 0.309 A and peak current at 0.5 A load is about 0.655 A — ample L1 margin. The datasheet specifies a 100 nF BST capacitor from BST to SW and recommends 2 × 22 µF output capacitance for 3.3 V.
Completed/fixed — aggregate 3V3 trunk: found a clear 2.29 mm bottleneck where the local 0.6 mm output network joined the 90.674 mm board trunk through two 0.2 mm segments. Those two segments were changed in place to 0.6 mm. The final whole-rail path from L1/output caps to the long trunk is 0.6 mm throughout; only the separate FB/load branch remains 0.2 mm.
Completed/fixed — C19: moved from (50.500, -24.800) mm, 270° to (48.075, -25.200) mm, 180°. Final pads: C19 BST at (48.850, -25.200) mm, C19 SW at (47.300, -25.200) mm. BST copper is a single 1.400 mm × 0.300 mm segment. U3 SW-to-C19 path is 2.000 mm total: 1.400 mm at 0.600 mm plus a 0.600 mm pad neck at 0.400 mm.
Completed — SW-to-L1: U3 SW (47.900, -23.800) → (47.900, -25.200) → (46.500, -26.600) → L1:1 (46.000, -28.500). Centerline length is about 5.345 mm, all 0.600 mm except the separate 0.400 mm C19 neck. There is no SW polygon or broad pour; switch-node copper is limited to the required traces.
Completed — output placement: L1 center (43.000, -28.500); output pad L1:2 (40.000, -28.500). C3 center (36.500, -30.200), 3V3 pad (37.450, -30.200); C12 center (37.500, -25.400), 3V3 pad (38.450, -25.400). Routed L1-output-to-cap path length is about 3.30 mm to C3 and 3.31 mm to C12, all 0.600 mm.
Completed — FB review: U3 FB at (48.850, -21.200) uses a 0.200 mm dedicated sense branch, transitions to Bottom at (48.830, -22.1215), and joins the quiet post-inductor 3V3 node near (39.250, -22.275) / the output-cap trunk, rather than the SW node. The Bottom route is separated from top-layer SW copper by the Mid-Layer-1 ground plane.
Completed — VIN loop review: C2 center (45.000, -22.500); VIN pad (45.000, -21.550) and GND pad (45.000, -23.450). U3 VIN is (46.950, -21.200) and GND is (46.950, -23.800). The local VIN entry uses a 0.500 mm capacitor neck followed by 0.200 mm pin necking; the direct C2-GND-to-U3-GND route is 0.500 mm. No focused clearance/overlap finding was reported.
Completed — unrelated disturbance review: historical placement diff for the prior executor’s rework only moved L1, C3, C12, C2 and C19; no unrelated component placement was changed. This independent repair changed only C19 and the two 3V3 bottleneck trace widths, plus local BUCK_BST/BUCK_SW traces.

Final 3V3 trunk measurements

Table

Segment	Layer	Width
L1:2 (40.000, -28.500) → (39.000, -28.500)	Top	0.600 mm
(39.000, -28.500) → (39.000, -26.000)	Top	0.600 mm
Output-cap branches from (39.000, -28.500)/(39.000, -26.000)	Top	0.600 mm
(39.000, -24.200) → (40.222, -23.247)	Bottom	0.600 mm
(40.222, -23.247) → (39.250, -22.275)	Bottom	0.600 mm
(39.250, -22.275) → (39.250, -22.9938)	Bottom	0.600 mm
(39.250, -22.9938) → (39.1085, -23.1353)	Bottom	0.600 mm
(39.1085, -23.1353) → (-51.5652, -23.1353)	Bottom	0.600 mm
Dedicated FB/load branch from (39.250, -22.275) toward U3 FB	Bottom/Top	0.200 mm

For the 90.674 mm long 0.6 mm segment alone, using 35 µm copper gives a conservative end-to-end resistance of about 74.4 mΩ, or 37.2 mV at 0.5 A / 27.2 mV at 0.365 A if the entire current traversed that full length. Actual drop is lower as loads branch along the trunk.

Fresh focused verification

Routing feasibility check: board valid, 60 components / 64 physical nets, no routing-feasibility anomalies.
Airwires: 0.
Redundant traces: 0 groups / 0 nodes.
Focused DRC at 2026-09-10T23:08:41Z: no findings for airwires, redundant traces, overlapping copper, overlapping traces, pad-access clearance, under-connected vias, dangling traces, floating copper, keepouts, or component-body overlap.
Automated Power Integrity category: the checker returned no available Power Integrity review category, so rail verification used exact copper geometry plus conservative DC resistance/current calculations.

Remaining issues / cautions

C3 and C12 retain the generic “Decoupling Loop Area” warning. This warning existed before and remained after an attempted local-via experiment; the experiment was removed because it did not clear the check and created redundant-trace candidates. The 3V3-side geometry is compact and correctly 0.6 mm, but the checker still considers the cap-to-U3 ground-return geometry non-ideal. A further fix would require a more invasive output-cap/L1/U3 placement and ground-return rework, not a clear local trace correction.
L1 nominal value is 3.3 µH rather than the table’s 3.9 µH recommendation, but it is within the datasheet’s stated usable range and has very large current margin; no change was made.
A pre-existing U1 board-edge/body-overhang warning exists outside this focused buck scope; it was not introduced or altered here.

Result summary
Final 3V3 trunk measurements
Fresh focused verification
Remaining issues / cautions