nRF52 periodically waking up, power consumption, zephyr, nrf connect sdk

[who9vy]

Currently, I'm facing a problem, that a custom board somewhat doesn't work. The main only consists of

#include <nrf.h>

void main(void) {
  while (1) {
    __WFE();
    __SEV();
    __WFE();
  }
}

The problem is, that the cpu seems to wake up every (about) 500us. The current measurement shows peaks every 500us with about 2mA and then back again to nearly 0.

If this is due to the custom hardware, I cannot see why because, there is nothing particular on the pcb. Which component could be the reason for the periodically wake up?

[qwertynoon]

Hi. Do you use RTT debugger? Maybe because of this

[who9vy]

Yes, but when measuring current, the debugger is not connected and all the logging is disabled.

Maybe, I'm doing something wrong. Can you give me a hint what to do?

[qwertynoon]

Actually if you use Zephyr then you don't need to include that code(__WFE... __SEV...)

Find attached folder "samples" in the root of Zephyr path and try to run some one.

What IDE do you use?

[who9vy]

Hi qwertynoon,

yes, before using __WFE, etc. I used the pm module of Zephyr but there hasn't been much difference.

Yesterday, I tried to use the NRF5 SDK v17.1 with the minimal code and the result is the same. Thus, I don't think that it is a problem of Zephyr anymore.

But if it is hardware, what can it be? I just used the NRF52810 reference design (DCDC) and added two connections to GPIOs. These I have removed from one PCB, but the result is the same.

Here is the Schematic of the PCB

I marked the parts which I removed from the PCB while measuring. Thus, I think this is the exact same as in the reference design but I don't get it below 250uA. And as seen in the image which shows the measurements the CPU seems to wake up periodically.

Have I missed something which should be different?

[Kenneth]

Considering you measure it both with nRF5 SDK and nRF Connect SDK, then it seems like some hardware issue. Do you disconnect the SWD interface (and possible other circuitry) and power cycle the nRF52810 before measuring current? Does running a simple toggle a GPIO work as intended?

Kenneth