Obtaining less then 50uA with 52840 dongle and PPKII

[androbi]

Hi,

after fixing a missing ground in my setup (see devzone.nordicsemi.com/.../interaction-jlink-and-power-profiler-ppkii) I have not been able to get less than 50uA of idle current on the 52840 dongle, when it should be around 2-3 uA. My main.c is

#include <zephyr.h>
#define SLEEP_TIME_MS   10000
void main(void)
{
	while (1) {
		k_msleep(SLEEP_TIME_MS);
	}
}

and I use CONFIG_SERIAL=n in prj.conf. I guess I have read almost all posts here on the forum about low power modes, but I could not find anything similar. I have tried the PPKII in source and in ampere meter mode (using a 3V battery), the USB is not connected and power is supplied using VDD Out pin with SB2 cut and SB1 soldered. I have measured with the debugger disconnected and done a power reset after disconnecting. All with the same result, around 50uA of mean current at 3V.

When using source mode on the PPKII I can drop the voltage to 1.9V and then I get around 15uA. Why should the current drop with the supplied voltage?

From the schematics the 52840 dongle does not have any passive elements which could source current (apart from the leds and , so I'd suspect power consumption coming from the SoC itself. But I have even measured 50uA @ 3V using

pm_power_state_force((struct pm_state_info){PM_STATE_SOFT_OFF, 0, 0});

which should be the lowest power mode possible. I am really stuck with this, any help appreciated!

[androbi]

Of course we can never 100% exclude the possibility of a wrong setup, but I have been using the unit very carefully. I usually only operate it in source mode or with a max. 4.2V Lipo in amperemeter mode, so a too high voltage can probably be excluded. As I commented above, the unit still works (sourcing and sampling) but not in the lowest uA scales, the failure of some of the components (op amps for low uA input?) could certainly also be a possibility. Would you like the photo with the device closed or opened so you can see the components? Please confirm in this case that opening does not void the warranty. Kind regards, Claus