1.0 PCB Self-Heating

[Poofjunior]

The PCB has ~3 degrees C of self-heating.

Putting the PCB in an enclosure will make the effect worse.

The primary source of heat is likely the LDO regulator. (These are inherently inefficient because they burn off excess voltage as heat.) The next suspects are the LEDs (LEDs were blinking in this case) and then the MCU.

There are a few options here:

1. Move the sensor to the corner of the board
2. Add cutouts to the ground plane around the sensor to reduce sensitivity
3. Switch the LDO regulator for a switching regulator.
4. Lower the clock frequency of the MCU
5. disable the LEDs during normal operation.

Note

It's worth confirming that the regulator is the problem using a thermal camera.

Regulator Option

The RT6150B is a good candidate for a new power supply option. It is 90% efficient in the operating region we will use, and it's identical to what's used on the Pico 2 board. See Pico 2 Datasheet. It will cost more board space, though.

[Poofjunior]

Update: Board has been redesigned with the following changes.

Top

Bottom

• added keepout regions for copper-poured features on all layers around the sensor to reduce heat transfer
• swapped linear regulator for a switching regulator that is 90% efficient
• moved the sensor to the corner of the board, further away from primary heat sources.
• increased PCB size to accommodate M3 mounting holes
• moved USB C connector and audio jack off of the board edge to account for case width.

[Poofjunior]

Final board has an exposed thermal pad on the bottom:

Worst case: we can also put a heatsink on the RP2040 directly:
Heatsink option