Ground plane discontinuity under switching regulators (2-layer board)
Nadie ha tomado este issue todavía.
Evaluación
- Dificultad
- 4/5
- Tiempo estimado
- 3-5 días
- Aptitud para principiantes
- 42/100
- Tipo de issue
- Error
- Claridad
- Bastante claro
- Estado de actividad
- Tranquilo
- Área
- embedded-iot
Línea de trabajo
Start by opening the Okapi PCB layout and inspecting the switching-loop areas around U1, U2, and U4. Check bottom-pour continuity, output-capacitor ground returns, exposed-pad stitching, and routing near sensitive analog signals. Done means each listed layout condition has been reviewed and corrected where needed, with any remaining respin or EMI concerns documented.
Escrito por el modelo de indexación a partir del texto del issue.
Descripción
Summary
The Okapi board is a 2-layer PCB. Switching regulators (LT3652 solar charger, TPS62822 step-down, TLV61220 boost) require a continuous, low-inductance ground plane under and around their switching nodes to control EMI and maintain loop stability. On a 2-layer board, the bottom copper layer typically serves as the ground plane, but routing traces through that layer creates interruptions that force return currents to take long, high-inductance paths.
Specific concerns
LT3652 (U4, solar charger):
- High switching currents (up to ~1.5 A) flow in the LT3652 switching loop: U4 SW pin → inductor → output cap → GND → U4 GND
- Any trace routing that interrupts the bottom copper under this loop increases parasitic inductance, increasing output voltage ripple and EMI
- The LT3652 evaluation board (DC1489A) uses an unbroken copper pour on the bottom layer directly under the IC and inductor
TPS62822 (U2, 3V3_AUX step-down):
- Same concern: the SW node, inductor, and output capacitor form a high-dI/dt loop that needs a solid ground reference
- TI layout guidelines for TPS6282x explicitly state: place input and output caps and the inductor in a tight cluster, with the exposed pad via stitched to the ground plane
TLV61220 (U1, boost converter):
- Boost converters are particularly sensitive to poor ground plane due to the high-frequency current pulses on the input side
- TI application note SLVA344 shows the critical ground pour must be uninterrupted
On a 2-layer board this is hard to fix without a respin
Unlike a 4-layer board where a dedicated internal ground plane is always available, a 2-layer board must use the bottom layer as ground — but this conflicts with routing needs. Options:
- Minimize top-layer crossing traces under regulators: Keep all signal routing away from regulator switching loop areas; use the bottom layer only for ground fill in those zones.
- Use copper pours with thermal reliefs disabled in the switcher areas on both layers; use stitching vias liberally.
- Route sensitive analog signals (ADC inputs, crystal, AREF) far from switcher switching nodes — especially the LT3652 SW node which sees the full battery/solar voltage.
- Add a common-mode ferrite on the 5V output of U3 (if applicable) if radiated EMI is measured to be a problem.
- Long-term: consider moving to a 4-layer stackup for future board revisions. Cost is modest at low volumes; the benefit in ground plane quality, EMI, and layout flexibility is significant.
Immediate action (current board revision)
Review the layout around each switcher's switching loop and confirm:
- The bottom copper pour is continuous (no routing traces crossing under the IC or inductor)
- Output capacitor GND returns connect directly to the ground pour, not via a long trace
- The exposed pad (EP) of each switching regulator is via-stitched to the bottom ground pour with ≥4 vias
🤖 Generated with Claude Code
- Lenguaje dominante
- G-code
- Estrellas
- 8
- Forks
- 2
- Métricas de merge de PR
- Sin PR fusionados en 30 d
Preparar el entorno
- Sin Dockerfile ni archivo de Docker Compose
- Sin plantilla de pull request
- Leer la guía de contribución
Primeros pasos
- Lee el issue completo y luego la guía de contribución del proyecto.
- Comenta en el issue que vas a ocuparte — evita que dos personas hagan lo mismo.
- Haz un fork del repositorio y trabaja en una rama.
- Abre un pull request que haga referencia al número del issue.
Más de NorthernWidget/Project-Okapi
-
bug
Dificultad 2/5 1-3 horas Aptitud para principiantes 72/100
-
IC1 AREF tied directly to AVCC (5V_AUX) with no bypass cap — internal reference conflict riskAbiertobug
Dificultad 2/5 1-3 horas Aptitud para principiantes 68/100
-
bug
Dificultad 5/5 Más de una semana Aptitud para principiantes 35/100
-
Re-land the charger's !FAULT trace on !SHDN so the firmware can detect a LiPo cell under sunAbiertoenhancement
Dificultad 4/5 3-5 días Aptitud para principiantes 48/100
-
enhancement
Dificultad 4/5 3-5 días Aptitud para principiantes 42/100
Todos los issues de NorthernWidget/Project-Okapi
Issues similares
-
bug
Dificultad 2/5 1-3 horas Aptitud para principiantes 75/100
Los mantenedores suelen responder en 1 día
-
bug
Dificultad 2/5 1-3 horas Aptitud para principiantes 78/100
iamankushpandit/Gume#231 ·
Los mantenedores suelen responder en 1 día
-
Dificultad 2/5 1-3 horas Aptitud para principiantes 76/100
DasBasti/IndiaNavi_Firmware#92 ·
Los mantenedores suelen responder en 1 día
-
[SoC]: Kirin 9010Abiertonew-soc
Dificultad 2/5 1-3 horas Aptitud para principiantes 68/100
hrw/arm-socs-table#231 · 2 comentarios ·
Los mantenedores suelen responder en 1 día
-
Dificultad 2/5 1-3 horas Aptitud para principiantes 82/100