nRF52840 Dongle draws more current on VDDH than VDD

[hestro92]

Hi,

We're currently developing a low-power framework for the nRF52840 and testing the application on a Dongle configuration. We have no problems when supplying the chip with a 3.3V voltage on the VDD pin to reach around 3.5-4 µA standby current. However when we change to supplying 5V on the VDDH pin instead we reach levels of 250 µA. We've made the appropiate cuts and connections concerning the SBs on the dongle and made no changes to choice of regulator. Do you have any idea what may cause this issue?

Thank you!

[Stian Røed Hafskjold]

Hi, what is the dongle version and the chip build code/revision?

[hestro92]

Hi, I believe the dongle version to be 1.2.0 and on the chip it reads N52840 QIAADO 1924AM, is that the build code/revision?

[Stian Røed Hafskjold]

Thanks. Earlier versions of the chip had some issues with the high voltage regulator, but the one you have is the newest and should be fine.

You say that you made the appropriate cuts and connections. When powering on VDDH you should not make any changes to the SBs. Power directly on the VBUS.

So SB1 open, SB2 closed. USB not plugged in. Power on VBUS.

If all of the above is correct, 250uA could be a pull resistor on a GPIO, or the HFXO crystal. Not sure what your code is doing but have you tried testing with just a __WFE() loop?

int main(void){

    for(;;)__WFE();

}

[hestro92]

Yeah sorry if I was slightly diffuse, we started out with supplying on the VBUS with 5V and made measurements around 250 µA, after that we decided to do the necessary modifications to run it with 3.3V on the VDD-pin just to compare and see if we got the same results; which in this case, we didn't.

The code runs so that it enters a WFE-loop and we get the proper behavior as we've seen with oscilloscope/amperemeter. Previously we've developed on the DK-version of the nRF and there we got 3.5-4 µA as well when supplying the board with external 3.3V which made us rather confused when we measured fifty times more current on 5V supply.

But shouldn't there be problems with both cases if a GPIO-pin pulls the current, why would it only affect the higher one? And the same for the crystal?

[hestro92]

Sorry I was wrong by a magnitude, we're getting 25-30 µA instead of 250 but the problem still remains that the current consumption is higher when supplying 5V. I also tried running your ble_app_blinky example app and removed code such that it only calls nrf_pwr_mgmt_run() thus only idling. The exact same issue arises here as well.

[Stian Røed Hafskjold]

Hi, I tried here as well and I see the same thing. I removed FB1 and the current dropped down to normal idle. So seems like D1 is consuming current when J1 is unplugged.