nRF52840 QFAA sensitiv to magnetic flux (custom board)

[uknoblic]

Dear Nordic team,

we are facing a strange phenomenon - in idle mode (k_sleep(K_FOREVER)) we notice, that the current consumption drastically increases if a magnet is placed close to the IC. Instead of a few mA, the current increases to ~10 mA. Any ideas where this might come from? No peripherals beside a few gpios + uart are used in our design.

[Kenneth]

Hello,

This is a first yes, do you see the same on the DK? I am thinking it's something else than the nRF52840 IC yes, possible the "can" of the 32MHz or 32kHz crystal that somehow magnetize and create a short of some kind. ~10mA is also a very strange current, the only thing close to this would be the radio actively transmitting or receiving, though I guess if you somehow stop the 32MHz it will also stop code execution. But it's slighty far fetched. If you can provide details on how to recreate this on a DK I am sure I will try to look into it more.

Kenneth

[uknoblic]

Good morning Kenneth, it'll need some days until the DK will arrive. I will come back to you.

[uknoblic]

By the way, meanwhile i figured out that the root cause for the increase current consumption seems to be the DCDC converter. When i turn it off via

  NRF_POWER->DCDCEN = 0;

i can't see any influence when i bring a magnet close to the PCB. The assumption is, that the chip inductors will saturate in this situation. I will double check this with a DK and come back to you