Testing power off current consumption

[nrbrook]

I'm trying to ensure my firmware goes into power off when it should, and that current consumption is as expected when in power off. However, using PPK2 to measure current on the nRF52 DK, I'm having difficulty getting a low power consumption even with an example project.

Is there an example which specifically demonstrates the lowest possible power consumption in power off? If I edit examples/peripheral/template_project/main.c and add NRF_POWER->SYSTEMOFF = 1; at the start of main, the current consumption is between 2.5 and 15ma. What else do I need to add to that example to get to the 0.3uA specified in the data sheet?

[nrbrook]

I've tested this example project: https://github.com/NordicPlayground/nrf51-powerdown-examples

and with USB disconnected and powered off, I get an average current of ~700uA on the nRF52 DK.

SB9 is open and so I've shorted the nRF current measurement header.

[Simonr]

Hi Nick

Most of our BLE examples in the SDK should support power OFF by the press of a button, but I would suggest checking out the Power Profiling Application, which is intended for doing power profiling of your application, and should be able to enter both system OFF mode and other low power modes without any issues. Please give it a run and let me know if you still see this high current consumption.

Best regards,

Simon

[nrbrook]

Hi Simon,

This doesn't make any difference, current consumption is still the same at power off. It seems unplugging USB makes a big difference in power consumption, but it still only goes to about 600uA. Has the nRF52 DK actually been tested with this example and PPK2 to go to <1uA? If so, perhaps my DK is faulty.

I tried adjusting APP_CFG_NON_CONN_ADV_TIMEOUT in the example to make it enter power off faster, but I see that definition is never actually used, only APP_ADV_DURATION is used, so that's a bug in the example.

[Simonr]

Hi

I discussed this with one of the PPK developers to get a better understanding of what's going on here. First off, the DKs are not optimal to do current consumption measurements on, as the onboard debugger also draws power. ~700uA seems like a leak between the nRF and on-board debugger (OBD). The OBD state when the USB is disconnected is floating, which is not great when you want an exact current consumption. When measuring over the External IN on the DK it will also power (and measure the current draw) of the OBD as well as the nRF chip. 

We recommend using SMU mode, with PPK2 [GND] to DK [External in -] and PPK [VOUT] to DK [nRF_VDD] which is the top pin on SB40 on the DK. That way, you'll power nRF_VDD directly, and bypass everything else. This way you should get the most accurate current consumption of just the nRF52.

Best regards,

Simon

[nrbrook]

Hi Simon, thanks.

I don't think SB40 exists on the nRF52 DK? I tried using [External in -] and VDD_nRF on the shield debug connector, but the result is the same.

Regardless, I fixed the problem with my code on my custom circuit so now I get <3uA on that board and don't really need to test this on the DK.

Thanks