NRF9160 in sleep mode over 1 mA current consumption

[MoritzR]

Hello Nordic,

we want to use the nrf9160 to measure some stuff and send data to a server while being battery powered. Therefore we need the lowest power consumption possible.

To get straight into the problem:I flashed it (via Bootloader) with a bare minimum of code to check and understand the current consumption, but eventough everything is deactived (especially Uart, which seem to be a common problem) it consumes about 1.27 mA. The nrf9160 is mounted on the Actinius Icarus board, nothing more connected to it.

Now you might think, the problem is the Icarus board. But if I set

NRF_REGULATORS->SYSTEMOFF = 1;

... the current will drop to 33 uA (This is still unexpected high, but the Icarus board features some devices which might be the reason for it)

So the question is, why does it consume so much power when being in idle mode?

main.c

#include <zephyr.h>
#include <sys/printk.h>

void main(void)
{
	k_sleep(K_SECONDS(20));
	NRF_REGULATORS->SYSTEMOFF = 1;
}

proj.conf

CONFIG_GPIO=n
CONFIG_LOG=n
CONFIG_CONSOLE=n
CONFIG_UART_CONSOLE=n
CONFIG_RTT_CONSOLE=n
CONFIG_USE_SEGGER_RTT=n
CONFIG_LOG_BACKEND_RTT=n
CONFIG_SERIAL=n
CONFIG_AT_HOST_LIBRARY=n
CONFIG_STDOUT_CONSOLE=n

# Zephyr Device Power Management
CONFIG_DEVICE_POWER_MANAGEMENT=y
CONFIG_PM=y

# Modem 
CONFIG_NRF_MODEM_LIB=y

#Boot
CONFIG_BOOTLOADER_MCUBOOT=y

also in the overlay file I set almost everything to deactivated:

&adc {
	status ="disabled";
};

&gpiote {
	status = "disabled";
};

&gpio0 {
	status = "disabled";
};

&uart0 {
    status = "disabled";   
};

&uart1 {
    status = "disabled";   
};

&uart2 {
    status = "disabled";   
};
&spi3 {
    status = "disabled";    
};

&i2c2 {
    status = "disabled";   
};

&timer0 {
	status = "disabled";   
};

&timer1 {
	status = "disabled";   
};

&timer2 {
	status = "disabled";   
};

&pwm0 {
    status = "disabled";   
};

Thanks for your help, I really appreciate it

Moritz

[Kazi Afroza Sultana]

Hello,

 Can you try to call this function lte_lc_power_off () from main function for turning off the modem? It could be the modem which draws current. 

This  shuts down the modem and puts the chip to system OFF mode ((+) Getting started with current measurements on the nRF9160 - Hardware design, RF test and coverage - Cellular IoT guides - Nordic DevZone (nordicsemi.com)). You can add this in your main function and measure the current again.

#include <modem/lte_lc.h>
void main(void)
{
lte_lc_power_off();
k_sleep(K_MSEC(1000));
NRF_REGULATORS->SYSTEMOFF = 1;
}

Best Regards,

Kazi Afroza Sultana

[MoritzR]

ok, I got it compiling now with the suggested lte_lc_power_off(). I had to change my proj.conf to the follwoing:

#Boot
CONFIG_BOOTLOADER_MCUBOOT=y

#Serial
CONFIG_SERIAL=n

# NEWLIB C
CONFIG_NEWLIB_LIBC=y

# Zephyr Device Power Management
CONFIG_DEVICE_POWER_MANAGEMENT=y
CONFIG_PM=y

# Modem 
CONFIG_NRF_MODEM_LIB=y
CONFIG_NET_SOCKETS=y

# Network
CONFIG_NETWORKING=y
CONFIG_NET_NATIVE=n
CONFIG_BOARD_SELECT_SIM_EXTERNAL=y


# LTE link control
CONFIG_LTE_LINK_CONTROL=y
CONFIG_LTE_NETWORK_MODE_LTE_M=y
CONFIG_LTE_AUTO_INIT_AND_CONNECT=n

But the current draw is still the same: 1.27 mA

Do you have any more ideas?

Thank you very much :)

[Kazi Afroza Sultana]

Hello,

I am thinking to look at your setup. You can send a picture of the measurement setup. I also would like to know how the setup is connected and what you are using to measure the current. 

Thanks. 

Best Regards,

Kazi Afroza Sultana

[MoritzR]

It is super basic. I attached pictures of two configurations I tried. The battery configuration seems to be the most efficient. When powering through the 5V port is seems to have a much higher Regulators_systemoff current, thats why I always took the battery configuration.

I am using a standard multimeter (Voltcraft VC185), it can measure down to 0,1 uA.

The fact that I can bring down the current to 33uA in battery configuration via software tells me, that there might be still a problem in software. But I am open for all ideas :)

[Kazi Afroza Sultana]

Hello,

The nRf9160 mounted on the board is seen in this picture is revision B0. This is old and will not work with the newer modem FW and SDK. There is difference how the modem is turned on and off between the B0 and the newest B1 version. You can get a newer version if your board has the B0 one.

Looking at the datasheet I think you can try to use VBAT as measurement place not USB.  

You can see USB is connected to FTDI chip, so its better not to power up that. 

Thanks.

Best Regards,

Kazi Afroza Sultana

[MoritzR]

That is a good hint! Thanks for spotting it. I indeed have a B0 version, the new version will arrive soon.

Nevertheless, I think the current is too high, even for the first version and also we dont want to throw the old version.

Which FW and SDK do you recommend for B0?

And how can I control the modem in version B0?

I will focus on measuring current through VBAT now.

Best regards,

Moritz

[Kazi Afroza Sultana]

Hello,

You can check the compatibility matrix here (https://infocenter.nordicsemi.com/topic/comp_matrix_nrf9160/COMP/nrf9160/nrf9160_modem_fw.html?cp=2_0_3_5 ). Any MFW v1.2.x (or older) must be used with B0 devices. 

You can find this from changelog also from download page (Compatible Downloads - nordicsemi.com) for each version of modem FW. For example: MFW 1.3.2 can be used for nRF9160 revision 1, but only for testing and development. 

Thanks.

Best Regards,

Kazi Afroza Sultana

[MoritzR]

Hello,

I received a board with version B2 on it. I updated the modem firmware via the "fmfu_smp_svr" example to v1.3.2, additonally I updated the nrf sdk to v2.0.2.

When in idle mode, with or without lte_lc_power_off(), it draws around 700 uA, which is much better, but still high.

I set CONFIG_SERIAL=n in proj.conf to disable the UART Module. Current is still at 700 uA and serial terminal only shows spm startup, which tells me some part of Serial is already disabled, but not everything.

I found that in \build\spm\zephyr\.config CONFIG_SERIAL is still set to "y" (I guess thats why it prints spm startup). In this thread they try to disable it. According to it, I added a file under \child_image\spm\proj.conf:

CONFIG_IS_SPM=y
CONFIG_FW_INFO=y
CONFIG_GPIO=y
CONFIG_SERIAL=n

But it wont build and give the following error:

warning: UART_CONSOLE (defined at drivers/console/Kconfig:43) was assigned the value 'y' but got the
value 'n'. Check these unsatisfied dependencies: CONSOLE (=n). See
http://docs.zephyrproject.org/latest/kconfig.html#CONFIG_UART_CONSOLE and/or look up UART_CONSOLE in
the menuconfig/guiconfig interface.

Can you help me how to disable the serial module properly? What is missing?

Thank you very much!

Moritz

[Kazi Afroza Sultana]

Hello Moritz,

It shows warning; so there should not be problem in building. Can you please send us the entire build log and also if possible the project file so we can look at and reproduce the issue from our side as well?

Thanks.

Best Regards,

Kazi Afroza Sultana

[MoritzR]

Ok, for those who are interested in this issue: I found a workaround: I am just setting the line "CONFIG_SERIAL=y" to "n" in build/spm/zephyr/.config and click on build. That works and reduces the power consumption.

Another thing I found why I was getting high current draw with the old version (B1) is, because one analog input pin was broken (no analog reading possible). I guess that was one of the main problems.