High current consumption (~20 mA) during button press on nRF54L15 DK (Matter Contact Sensor sample)

[IOANNIS0]

Hello Support Team,

I am currently working with the Matter contact sensor sample (ncs\v3.4.0\nrf\samples\matter\contact_sensor) on the nRF54L15 Development Kit (DK).

As part of my application requirements, users need to keep two or more buttons depressed for extended periods. However, when pressing the SW1 button on the DK, the Power Profiler Kit 2 (PPK2) registers a constant high current draw of approximately 20 mA. This current consumption significantly exceeds what would be expected from internal pull resistors.

The board is powered via the onboard nPM1300 PMIC. Despite testing various configurations—including both internal and external resistor setups,using sense-mask,using gpio_pin_configure_dt —the current spike persists consistently as long as the button is held down.

Since the final product will be battery-powered, this level of power consumption is a critical issue for the design. Could you please advise on potential solutions to resolve this behavior, or let me know if this indicates a hardware defect on the nRF54L15 DK?

Thank you for your assistance.

.

[Carl Richard]

Hello!

Thanks for reaching out.

From what I can see in your screenshot the average current during the button press is quite far below 20 mA, with some peaks at around 15 mA. With that said, the current draw seems quite high for only a button press. Does anything else happen when the button is pressed? Also, try to capture the average current during the actual button press as well.

Best regards,
Carl Richard

[IOANNIS0]

Dear Richard,

Thank you for your prompt response.

The program is currently functioning as intended; the input successfully toggles within Home Assistant, and all device details are correctly populated.

However, we have identified only this significant drawback: high current spikes occur whenever the button is held down. Debugging logs—captured during active operation, as the issue is intermittent—confirm that the program correctly enters the interrupt only once on edge changes, as expected.

In summary, while the core logic and operations are running correctly, resolving these current spikes remains the primary issue to address.

Best regards,

IOANNIS0