app

Adom Fields

Public Made by Adom

by adom

Adom Fields: where the current and the heat go on your board.

kicadpcbcurrent densitythermalheatcopper poursvias3danalysisedacurrent flow animationvoltageriver
Overview Files Skills Discover Packages Releases Dependencies Issues3 Breadcrumbs Pull Requests Activity
Closed PR #13 released in 0.10.6 on 5 Nov 2026 bug report

Decoupling check expects VCC2 decoupled to AGND; LM251772 decouples VCC2 to PGND, giving false decoupling-missing findings

Glenn Jones · ·closed by Colby Knox

We kept a list of tool issues as we worked on a project, the 772 power module (LM251772), a 4-layer board built with AI Flow on the Fusion lane, and set them aside so they wouldn't slow the build down. Now that the project is finished, we're filing them together. Some were seen on earlier versions; where we could, we checked that they still apply.

Versions: adom-fields 0.10.5.

What happens: on the 772 power module (LM251772 4-switch buck-boost), the review reports 4 decoupling-missing findings for VCC2 because it looks for capacitors from VCC2 to AGND. The LM251772 datasheet decouples VCC2 (the gate-drive supply) to PGND, which is how C16/C17 are connected.

Expected: a gate-drive supply decoupled to the power ground counts as decoupled (or the spec can name the return net per rail).

Related but different: #25 (which net is taken as the board ground).

No design files attached; happy to describe more on request.

1 Reply

Colby Knox marked this in progress ·
Colby Knox mentioned this in pull request #13 ·
Colby Knox ·

Fixed in adom-fields 0.10.6. Upgrade with adom-wiki pkg install adom/adom-fields.

What changed

  • A decoupling capacitor now counts when it returns to any ground the chip has a pin on, or to the review's ground as before. So C16 and C17 from VCC2 to PGND count, as long as the LM251772 has a pin on PGND.
  • review.decoupling now records the return net for each pin.
  • You don't need to name the return net in the spec, so I didn't add a key for it.

What is still reported, on purpose

  • A power pin with no capacitor.
  • A capacitor to a ground the chip has no pin on.

The finding now names every ground that would count, and suggests a capacitor to the ground of the chip's nearest ground pin.

How it was tested

  • A test board with a chip on AGND and PGND: VCC2 to PGND and VCC1 to AGND are no longer findings. A rail with no capacitor and a rail whose capacitor goes to a third ground still are.
  • A real Fusion board with AGND and PGND (a BQ25792 charger): two false decoupling-missing findings are gone. They were on LDO_3V3 at chips with only PGND pins, with their capacitors to PGND.
  • A one-ground 12 V buck: every finding unchanged.

Not tested: your 772 module itself, since we don't have the file. If VCC2 still shows up there, please tell me which nets its pins and capacitors are on.

Still open: which net counts as the board's ground is unchanged and tracked in #25, and --with pdn still looks for a rail's capacitors on that one ground. So a rail decoupled only to PGND on a board whose ground is AGND gets no impedance verdict from it.

Colby Knox closed this ·

Log in to reply.