1mA power draw from nRF52840 when asleep, but not off

[Gordon]

Hi!

I've been developing software for nRF52832 devices for a while now, and it's been pretty easy to get into low power modes.

However using the exact same software, but built for nRF52840 I'm seeing a very surprising constant 1mA power draw even when supposedly asleep.

The same 1mA power draw occurs on 3 completely different types of nRF52840 device (with or without USB), and when I do `sd_power_system_off` it goes away, so I'm pretty sure it's the chip.

I've tried to check the usual stuff (GPIOs set wrong, UART/PWM enabled) but it seems it's not those.

Any ideas what I can check to try and debug this? Are there any common gotchas related to the DCDC and regulator setup vs the nRF52832?

The strange thing is it's pretty much the exact same power draw in every case - I'd say maybe 0.95mA above what I'd expect it to be.

Thanks!

[Stian Røed Hafskjold]

Hi, First thing that comes to mind is errata 195. It was introduced on the nRF52840, so it's not present on the 832. It also fits well with the 0.95 mA current draw. It would also go away when entering system OFF.

Please have a look here and check if it could be relevant:

https://infocenter.nordicsemi.com/topic/errata_nRF52840_Rev2/ERR/nRF52840/Rev2/latest/anomaly_840_195.html?cp=4_0_1_0_1_26

The regulator design is pretty much the same as on 832, except if you are using the high voltage (VDDH) input. But it should not cause 1mA of current draw anyways.

[Gordon]

Thanks - that's great news! Just gave this a try and it shaved a good 700uA off the power draw. I hadn't previously bothered turning off peripherals like SPI when the device sleeps - should I? Or in this case is it just because of that errata?

[Stian Røed Hafskjold]

Some peripherals will draw current in the 10-20 uA range when not completely disabled. This includes slave peripherals like the SPIS, TWIS and UARTE. The SPIM should not consume any extra current when not transmitting. In this case it's only because of the errata.