Not able to achieve minimum sleep current on NRF54L15

[payalD]

Hello Team,

I am trying to measure the sleep current of NRF54L15 DK, in the system ON mode, for that I am using Blinky sample code, from SDK version 2.9, which uses the k_msleep() function. Using this function, I toggle the LED every 1 minute (1 min ON, 1 min OFF). Sleep current measured using this sample is 584 micro Amps when the LED is ON and 580 micro Amps when the LED is OFF, which is far more than expected. Can you please help me on how to reduce the sleep current? Is there any other example code for sleep current measurement in the system ON mode?

Thanks,

Payal

[Vidar Berg]

Hello Payal,

Have you disabled serial logging over UART (CONFIG_SERIAL=y)? If not, then that may increase the high idle current since the UART will request the HF clock to be kept active in sleep. You may also try this sample:

 current_meas_test.zip

Best regards,

Vidar

[payalD]

Hello Vidar,

Thanks for your quick reply.

I tried the sample code provided by you, and the following are the sleep current results:

1. All RAM Enabled: 4 micro Amps

2. 128k RAM Enabled: 3.1 micro Amps

3. 64k RAM Enabled: 2.7 micro Amps

Are these the expected results?

If yes, I have a follow-up query. When I want to keep all RAM Enabled, and want to configure 1 GPIO as interrupt mode, while the device is in system-on sleep mode, for wake-up. When I tried to configure the GPIO as interrupt sleep current increased to 22 micro Amps. Is there any way to reduce the sleep current in this scenario?

[Vidar Berg]

Hello,

The measurements appear to be a bit on the high side. Could be chip to chip variation (within normal distribution) or leakage on the board. You can try to disconnect the connections mentioned below via the board configurator in nRF connect for desktop to see if it helps.

https://docs.nordicsemi.com/bundle/ug_nrf54l15_dk/page/UG/nRF54L15_DK/measurements/set_up_dk.html 

payalD said:

When I tried to configure the GPIO as interrupt sleep current increased to 22 micro Amps. Is there any way to reduce the sleep current in this scenario?

Which Port did you use? You should also use the Pin sense mechanism for low power.

https://docs.nordicsemi.com/bundle/ps_nrf54L15/page/gpio.html 

&gpio0 {
	/* Use PORT event rather than GPIOTE IN event, to save power */
	sense-edge-mask = <0xffffffff>;
};

[payalD]

Hello Vidar,

Vidar Berg said:

The measurements appear to be a bit on the high side

What is the expected sleep current in all RAM-enabled mode?

Vidar Berg said:

You should also use the Pin sense mechanism for low power.

The Pin sense mechanism allowed me to reduce the sleep current to 4 micro Amps while detecting interrupt in sleep mode.

[Vidar Berg]

Hello,

Typical sleep current for some selected configurations are documented in the Current consumption section of the datasheet.