NRF5340DK Power consumption while in HV Mode

[pauleffect]

Hello.

Im using a 5340DK to measure its current consumption while in HV mode.

NCS version is 2.4.1   The .conf file below.

The board is supplied via VDD_BAT and there are two external components (2), supplied from the VDD pin of the board (accurately measured while the firmware is running). 

# RTOS
CONFIG_MAIN_STACK_SIZE=4096
CONFIG_HEAP_MEM_POOL_SIZE=4096
CONFIG_EVENTS=y

# POWER MANAGEMENT
CONFIG_ADC=y
CONFIG_ADC_NRFX_SAADC=y
CONFIG_NRFX_SAADC=y
CONFIG_PM=y
CONFIG_PM_DEVICE=y
CONFIG_NRFX_POWER=y
CONFIG_BOARD_ENABLE_DCDC_APP=y
CONFIG_BOARD_ENABLE_DCDC_NET=y
CONFIG_BOARD_ENABLE_DCDC_HV=y

# SECURITY
CONFIG_NRF_SECURITY=y
CONFIG_MBEDTLS_PSA_CRYPTO_C=y
CONFIG_MBEDTLS_ENABLE_HEAP=y
CONFIG_MBEDTLS_HEAP_SIZE=8192
CONFIG_MPU_ALLOW_FLASH_WRITE=y
CONFIG_FPU=y

# MESSAGING
CONFIG_IPC_SERVICE=y
CONFIG_IPC_SERVICE_LOG_LEVEL_INF=n
CONFIG_IPC_SERVICE_BACKEND_ICMSG=y
CONFIG_IPC_SERVICE_BACKEND_RPMSG=n
CONFIG_MBOX=y

# NAND
CONFIG_SPI=y
CONFIG_SPI_ASYNC=y

# AUDIO
CONFIG_I2S=y
CONFIG_I2C=y
CONFIG_NRFX_I2S=y


CONFIG_BOARD_ENABLE_CPUNET=y

Im using SAADC, I2C, I2S (nrfx drivers), UART, SPI. The NET firmware sits idle. The idle power consumption of the board sits comfortably down in the uA range.

UART state is toggled in firmware, being enabled on demand when a line should be sent to pc.

I2C only enabled when needed.

SAADC is used for vddh voltage measurement @ 1hz frequency.

I2S is used in master mode @32Khz, 16bit, mono. The MCLK pin divisor is bypassed (12.288 Mhz output). Tx pin disabled.

SPI is only enabled when needed (roughly twice / minute, for ~ 300 ms).

---

With the device doing active work, the power consumption of the board is ~5mA when nothing is connected to the usb port of the nrf (1) and ~ 3.2 mA when I plug a usb cable into the aforementioned port. I supplied power via the USB port because I read here, on this forum (can't find the link), that some components on the dev kit would still need power (the switches, passives etc), so the board should still be supplied with power, not just the MCU part.

By active I mean I2S + external codec running.

Judging from the peripherals I use + the external components in use (which I have accurately measured with tools way more expensive than my car), I should be looking at a power consumption of less than 2.5 mA, mostly coming from the i2s peripheral.

Is the first measured figure (basically with the board only powered by VDD_BAT), 5mAm, correct or is it time to pop a champagne? 

EDIT: Clarifications added.

[pauleffect]

Thats what i was hoping for. But the leds are not powered and neither is the debugger. Isnt that the whole point of the nrf only switch?

[JONATHAN LL]

What USB connector are you using? 

And what position is SW9 in? (USB\VDD\Li-po)

Can you share the hex file so that I can try and reproduce the issue?

What version of the nRF5340DK do you have?

You don't need to flip SW6 to nRF ONLY to measure current used by the nRF. 

Regards,
Joanthan

[pauleffect]

I'm sorry, sir. I think I understand your confusion now. The first thing in my question is a picture of a DK. With every switch labeled with the position they're in. Clearly, you didn't see that. Probably some forum error. Anyway:

SW9: LIPO.

SW10: ON

SW8: ON

SW6: NRF ONLY.

J3 connected to a pc: ~ 3mA. J3 left disconnected: 5mA. The exact same piece of code.

The thing is, the 3.2 power consumption is what I estimated from the docs (and adding the obvious FUBAR coefficients, laws of murphy be observed). The 5mA figure is ridiculous though. I'm just trying to be as sure as humanly possible that the design is ok before proceeding, as I'm sure you understand.

EDIT: Just to be clear, I dont want to measure the MCU power consumption. I want to measure the power consumption of the MCU + the two external components while powered via NRF's VDD. Because the efficiency of the DC-DC isn't quite well documented (or I haven't been able to find info, on the forum or otherwise). I'm interested in profiling it, so I will make the decision to either include a PMIC (probaby nPM 1300) or run the MCU in HV mode directly. I would gladly accept, for example, a 500 uA hit in run mode, if that meant less components on the board.