RE: nRF52832 current consumption issues

Einar Thorsrud — Mon, 11 Feb 2019 10:00:07 GMT

Hi,

Nguyen Hoan Hoang 's answer is important. I have some additional comments (linked to your numbered questions):

1) The CPU frequency of the nRF52832 is always 64 MHz, so your calculation is not correct. It is essentially the same number as IDDFLASHCACHEDCDC (3.7 mA).

2) The current consumption numbers given for the SAADC (and other peripheral) only reflect the current consumption of the peripheral itself, not any resources it depends on. So the IADC,CONV number of 700 μA does not include the current consumption of the clocks, regulator or DMA logic. Therefore, the total current consumption is significantly higher. The system will consume significant amounts of power between sampling as well, as these resources are constantly enabled while the SAADC peripheral is enabled.

3) The best way to reduce the SAADC current consumption is to minimum between sampling is to disable the SAADC peripheral, provided that you do not need to use it frequently or have a very accurate sampling rate.

RE: nRF52832 current consumption issues

Nguyen Hoan Hoang — Sun, 10 Feb 2019 15:10:33 GMT

First thing before you do any current consumption test is to put the mcu to sleep without doing or activating anything else. This means in the main() just do a while (1) { __WFE(); }, nothing else. Compile in release mode. Then measure how much current, (no debugger or jtag connected). The current should be less than 10uA. Anything higher than that means you have hardware leaks in your design. Then you active one by all your io to find out where which i/o leaks more.

This example here https://github.com/I-SYST/EHAL/blob/master/ARM/Nordic/exemples/TPHSensorTag.cpp reads temperature, humidity, pressure and gas from the BME680, ADC to read battery level and advertises them over BLE. The average current is around 15uA.