This post is older than 2 years and might not be relevant anymore
More Info: Consider searching for newer posts

High Current Consumption in Thread/CLI example

Hi,

I flashed example thread\cli\ftd\usb\pca10056\blank\ses contained into nRF5_SDK_for_Thread_and_Zigbee_v4.0.0_dc7186b and I measured current consumption by using Power Profiler.
I followed all of steps written here https://infocenter.nordicsemi.com/index.jsp?topic=%2Fsdk_tz_v4.0.0%2Fthread_cli_example.html&cp=7_7_0_2_10_4_1 and I do the same for the dongle (but using PCA10059 FW).
When IPv6 is started (through CLI command ifconfig up) current consumption reaches 12mA avarage and when thread is started current peaks are about 18mA. (see below pictures)

Here IPv6 is started

Here below Thread is Started.

Am I doing right or this consumption is not regular?

Thanks a lot,

Simone.

  • Hi,

    I have not measured current consumption for CLI example, but I have done for NCP, which shows nearly exactly the same numbers. The radio will be in receive mode most of the time when the Thread stack is operating in the Router/Leader role, giving you the current consumptions shown in the product specifications.

    You should be able to reduce the currents quite a lot by enabling the DCDC regulator, if running on a DK or a board with the required components. Add this line to the start of main() in the example code:

    Fullscreen
    XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

    Best regards,
    Jørgen

  • Hi Jorgen,

    thanks for your response. 
    Well, I also measured current consumption when DK is operating in Child role and result is the same.
    I tried to fill in NRF_POWER->DCDCEN = 1 as you suggested but nothing changed.
    Do you have an idea how to set DK as Sleepy End Node (SED)?

    Thanks a lot,

    Simone.

  • The CLI example may still operate as a Full Thread Device (FTD) with radio in RX while idle, even if it is in the child state. You can try to change the mode using the CLI command "mode". I do not have a PPK with me at the moment to test with, but I think "mode n" should lower the current consumption. Anyway, the CLI example is not optimized for current consumption measurements, and it may use additional peripherals that increase current consumption (UART, HFCLK, etc). I would recommend you to test with one of the Sleepy End Device examples in the latest SDK release:

    Best regards,
    Jørgen