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

Where's the leak?

Hi,

I've been though some power profiling investigation of my application and incorporated the solutions discussed and here and the hardware solution discussed here but I am using a 1Meg Ohm resistor between the VDD rail and the GPIO PIN. The PIN is the n tied to the ground through a switch.

My development stack is

- An nRF52832 DK
- SDK 15.3.0
- Segger IDE
- The soft device is S112
- JLinkRTTViewer for debug
- I'm using the ble_app_template as my starting point having added the battery service and the link loss service


The power profiler kit that I used from Nordic returned some promising results having worked my way through the instructions below 

https://infocenter.nordicsemi.com/index.jsp?topic=%2Fug_nrf52832_dk%2FUG%2Fnrf52_DK%2Fhw_meas_current.html

and support from this board.

Results:

Advertising with PIN 12 pulled to ground

Average whilst in system off was promising as well

and the connection event

My application is quite elementary as it simply monitors a GPIO port and the battery.

  • Once a switch on GPIO PIN 12 is pulled to ground it pulls the chip out of sleep and advertises for 30 seconds with a 2 second advertising interval. It then sleeps for 30 seconds and repeat. It is unlikely that the switch will be closed for more than one hour a day.
  • I have only two services
    • link loss service
    • battery service.
  • Battery service is set to check the battery level immediately, once "Notify" is initiated by Central, and every 1 hour thereafter.
  • If the switch on PIN 12 is opened then the uC goes to sleep.

So getting this far with the power profiling I decided to put a fresh CR 2032 coin cell in my second unit (SB9 not cut) the DK board drained the battery  from 100% to 57% overnight to 42% over 24 hours with a maximum of 1 hour of advertising and about 3 connection events. 

The power supply information below insists that you never place the USB and coin cell in the system at the same time which I have respected although the diodes should have protects things as there are no solder jumpers on the board at SB11 or SB10.

I also note that my GPIO pin is configured to use GPIOTE PORT event with a NOPULL. As I have a 1M ohm resistor pulling it up to VDD I'm not doing anything silly there am I as per this post?

Questions:

  • Is the answer that I should not worry at this point about the battery behaviour of the DK and trust the Power Profiler kit results as I move on to create my custom board?
  • Is it because the DK is running a bunch of other stuff on the development board that makes the test obsolete as is discussed here? i.e. for a proper test I should cut solder bridge 29 or wait until my custom board is finished?

Thanks in advance and apologies if a silly question.

Best,

David

Related