Hi,
We developed a custom board based on the nRF54L15 SoC. We realized many current consumption tests with regulator configured in DC/DC mode. These measurements are in line with the SoC datasheet. Then we decided to switch the regulator in LDO mode since this configuration will be used in the final version as we cannot embed large inductors in our design. At this point we noticed that the current consumption is really high during sleep mode.
In order to confirm this, I did the same comparative on a nRF54L15DK evaluation kit. To realize good measurement in low power mode, I performed the followings steps :
In order to confirm this, I did the same comparative on a nRF54L15DK evaluation kit. To realize good measurement in low power mode, I performed the followings steps :
-
First I disabled the VCOM0, VCOM1 and LEDs power using the nRF Connect Board Configurator application.
-
I set up the DK, accordingly to the user guide, using a Power Profiler kit configured in ampere meter.
-
For the firmware part, I started from the code example provided in the Lesson 2 - Exercise 2 of the Bluetooth Low Energy Fundamentals course that starts a simple advertising process. I just removed all the part relative to the management of LEDs and button.
-
I updated the prj.conf file to disable the console and the log trace :
CONFIG_LOG=n
CONFIG_CONSOLE=n
CONFIG_UART_CONSOLE=n
CONFIG_RTT_CONSOLE=n
-
Finally I disabled useless peripherals by applying a device tree overlay to the default setup :
&uart20 {
status = "disabled";
};
&spi00 {
status = "disabled";
};
In this case, the SoC still configured in DCDC mode. As depicted by the following current profile, the consumption in sleep mode (measured between 2 advertising sequences) fits with the datasheet {~3 uA).
Then I add the following lines in the device tree overlay to switch the regulator in LDO mode :
(by the way, it seems that no dedicated Kconfig setting is required now for nRF54L15 SoC to switch the regulator mode. But I am not 100% sure)
&vregmain {
status = "okay";
regulator-initial-mode = <NRF5X_REG_MODE_LDO>;
};
In this condition, we can observe the current consumption during the sleep mode is quite high (> 90 uA), while I was expecting something similar !
I would like to understand this is due to a bad project setting, or if I missed something important in the nRF54L15 datasheet or errata concerning the current consumption in LDO mode ?
Strangely, it appears that the Online Power Profiler tool provided on the Nordic DevZone avoids the user to deselect the DCDC regulator option to perform simulations, unlike the others nRF SoC !