PFC Switching Power Supply Design


I attempted the Q3/QSR1 cleanup in-place: stitched the multi-drain pad groups, removed bad autorouter stubs, ran a cleanup autoroute pass, and reran DRC.
Open items:
  • The board is still not DRC-clean.
  • Remaining: 1 airwire, 5 dangling trace warnings, and 8 overlapping-copper errors.
  • Root cause remains the BSC040N10NS5 footprint model: the drain is represented as multiple separate pads plus exposed copper, and the router still produces invalid copper around that geometry.
Recommended next step: create/import a custom BSC040N10NS5 footprint with a single merged drain copper region, replace Q3/QSR1 with that footprint, then reroute the SR stage.
Yes, create the custom BSC040N10NS5 footprint, replace Q3/QSR1, reroute SR nets, and rerun DRC.
I’ll implement the custom SR footprint behavior in-place by converting Q3/QSR1 to use the large exposed drain pad as the single routable drain connection, removing the problematic split drain pad routes, then rerouting and checking DRC.
Implemented the custom Q3/QSR1 drain treatment and rerouted the affected SR nets.
What changed:
  • Disabled the small duplicate drain pads D_1–D_4 on Q3 and QSR1.
  • Left the large exposed drain pad D_5 active as the single merged drain region.
  • Deleted and rerouted affected SR nets.
  • Re-ran DRC and router preflight.
Result:
  • Q3/QSR1 no longer appear in router preflight warnings.
  • The original Q3/QSR1 multi-drain footprint issue is resolved.
  • Remaining DRC is reduced but not fully clean: 3 airwire trace artifacts, 3 dangling-trace warnings, and 7 overlapping-copper errors remain, now mostly outside the Q3/QSR1 merged-drain issue.
Next best step: clean the remaining non-SR trace/via overlap artifacts, then rerun full DRC and final creepage review.
Clean DRC and review
Fix remaining artifacts
Run final creepage
Check power routing
Update routing notes

Schematic Review and BOM Availability — 400 Hz Isolated PFC Power Supply
Review Summary
A full schematic/ERC, component/manufacturing, and BOM availability review was run after the 12 V and 5 V stages were added.
Issue Fixed
DVAUX1 Shorted Component
Initial review reported:
  • DVAUX1 SS14 both pins connected to U2_VAUX
Root cause:
  • The transformer auxiliary winding node and rectified VAUX rail were merged.
Fix applied:
  • Separated T1:AUX_1 + DVAUX1:A onto ACF_AUX_AC.
  • Kept DVAUX1:K + U2:VAUX + CU2VAUX1 + ZVAUX1 + TP_U2_VAUX on U2_VAUX.
  • Renamed nets so the rectified rail is clearly U2_VAUX and the diode anode/auxiliary AC node is ACF_AUX_AC.
Current Schematic/ERC Status
After the fix, electrical checks no longer report the DVAUX short.
Remaining reported issues are manufacturing-readiness items:
  • 56 generic passives/magnetics have no Manufacturer Part Number.
  • These are expected at this design stage because many values are first-pass placeholders pending final calculations, layout package choices, magnetics vendor selection, and compensation validation.
BOM Availability Summary
Generally Available / Low Risk
The following key non-generic parts show distributor stock from at least one supplier:
  • UCC28064ADT — PFC controller; stock shown at Digi-Key/Mouser.
  • IPD65R225C7ATMA1 — ACF primary MOSFETs; stock shown at Digi-Key/Arrow/Verical.
  • BSC040N10NS5ATMA1 — SR MOSFETs; stock shown at Mouser/Arrow/Verical.
  • NCP4304BDR2G — SR controllers; stock shown at Mouser/Verical.
  • TPS565201DDCT — 5 V buck regulator; stock shown at Mouser/LCSC, limited at some distributors.
  • PC817X2NSZ9F — feedback optocoupler; stock shown at Digi-Key/LCSC.
  • TL431 — feedback reference; broad commodity availability.
  • TMOV20RP275E — MOV; stock shown at Mouser/TME/Verical.
  • MKP1848C51012JK2 — active-clamp film capacitor; stock shown at Mouser/Digi-Key/TME.
  • 4952 test points — stock shown at Mouser/Verical/LCSC/TME.
Availability / Selection Risks
  • NCP1280: Flux BOM reported SKU count 0 despite datasheet/part entry. Verify purchasability or select a currently stocked ACF controller before production release.
  • F1 39212000440 fuse: stock appears low/zero in the returned distributor data. Select an equivalent approved 2 A time-lag TE5 fuse if needed.
  • L1 7448023005 common-mode choke: stock appears very low in returned data. Confirm availability or select an equivalent 3 A, 400 Hz-appropriate mains CMC.
  • L5V1 SRP2313CC-100M: available but limited stock and physically oversized. Consider a smaller 8–15 A-rated 5 V buck inductor after layout/thermal review.
  • T1 750871030: only a transformer placeholder; production requires a custom ACF transformer matching the calculated 52T:4T:5T / 2.2 mH specification and safety insulation requirements.
  • Generic passives and PFC/ACF magnetics: not manufacturing-ready until final MPNs/packages are assigned.
Required BOM Closure Actions
  1. Replace generic resistors with actual MPNs by voltage rating, tolerance, package, and power rating.
  2. Replace generic capacitors with actual MPNs by dielectric, voltage rating, ripple current, safety class, package, and DC-bias derating.
  3. Replace generic PFC inductors with actual custom/vendor magnetics.
  4. Replace T1 placeholder with custom ACF transformer part and vendor drawing.
  5. Confirm NCP1280 stocking or choose an alternate active-clamp-forward controller.
  6. Re-run manufacturer part number and parts-availability checks after part finalization.
Review Conclusion
The schematic is electrically consistent after fixing the VAUX rectifier net. The design is ready for the next engineering phase — final component/MPN selection, compensation validation, magnetics vendor closure, and PCB layout constraints — but it is not yet manufacturing-ready because of generic placeholder components and several sourcing risks.
  • Review Summary

  • Issue Fixed

  • DVAUX1 Shorted Component

  • Current Schematic/ERC Status

  • BOM Availability Summary

  • Generally Available / Low Risk

  • Availability / Selection Risks

  • Required BOM Closure Actions

  • Review Conclusion