SYSTEMOFF=1 doesn't work on nRF51422 after programming unless IC is power cycled

[sjacobsen]

We're using systemoff=1 to turn our system off, but the command doesn't work after programming, i.e. device immediately turns back on with 16 MHz RC osc running, and current consumption is approx. 4.0 mA.  If we cycle power after programming, the command does work effectively.

Here's our setup:

nRF51422 QFAAE0/F0

Segger programmer attached or detached from IC

Keil or nRFgo Studio

S210 v4.0.1 soft device

Here's the code context:

// ----------------------------------------------------------------------------------
void
SetSystemOff( // Put the beacon into system off mode (suspend)
void)
{
uint32_t ulReturnValue = 0;
uint32_t ulPowerdownTimeCount = 0;

// Ensure system timer is stopped
TimerStop(SYSTEM_TIMER_INDEX);

ulPowerdownTimeCount = 0;

// Disable all events from input GPIO
GPIOTEDisableInputPin(GPIO_PIN_VOLTAGE_SUPERVISOR);
GPIOTEDisableInputPin(GPIO_PIN_BUTTON_INPUT);

// Set the button to wake system on high signal
ulReturnValue = GPIOTEConfigureInputPin(GPIO_PIN_BUTTON_INPUT, GENERATE_EVENT_HIGH);
ErrorCheck(ulReturnValue);

// Wait for 500 milliseconds to ensure we don't suspend too soon after power on, and to debounce button signal
while(ulPowerdownTimeCount < BEACON_POWERDOWN_DELAY_100MS*5)
{
//Delay 50 ms
DelayUS(50*1000);
WatchDogReload();
ProcessGPIOEvents();
DebounceProcessing();

ulPowerdownTimeCount += 50;
}

// Stop timer
TimerStop(SYSTEM_TIMER_INDEX);

DisableSoftDevice();

// Set the power off flag if it's not set already
if(!gulPowerOffFlag)
{
ulReturnValue = NvmcPageErase(BEACON_POWERSTATE_FLASH_ADDRESS);
ErrorCheck(ulReturnValue);

WatchDogReload();

ulReturnValue = NvmcWriteWord(BEACON_POWERSTATE_FLASH_ADDRESS,1);
ErrorCheck(ulReturnValue);
}

// Set the general purpose register to indicate manual power down, we need this as a workaround to PAN #14
NRF_POWER->GPREGRET = ((BEACON_OFF_INDICATOR & POWER_GPREGRET_GPREGRET_Msk) << POWER_GPREGRET_GPREGRET_Pos);

// Go to system off
NRF_POWER->SYSTEMOFF = 1;

// Wait for suspend to occurr
while(1)
{}

}

// ----------------------------------------------------------------------------------
void
DisableSoftDevice() //If enabled disable the softdevice
{
uint8_t ucSoftDeviceEnabled = 0;
uint32_t ulReturnValue = 0;

ulReturnValue = sd_softdevice_is_enabled(&ucSoftDeviceEnabled);
ErrorCheck(ulReturnValue);

// If soft device is enabled, shut it down
if(ucSoftDeviceEnabled)
{
WatchDogReload();

// Stop all broadcast channels
ChannelStop(BROADCAST_BEACON_CHANNEL1);
ChannelStop(BROADCAST_BEACON_CHANNEL2);
ChannelStop(BROADCAST_BEACON_CHANNEL3);

WatchDogReload();

// Disable softdevice
ulReturnValue = sd_softdevice_disable();
ErrorCheck(ulReturnValue);
}
}

[Jørgen Holmefjord]

Hi,

This is expected behavior. When you have programmed the chip, it will remain in debug mode until a reset is performed. When the chip is in debug mode, you will enter "emulated system off mode", where debugging capabilities are retained. Please have a look at page 44 of nRF51 Reference Manual v3.0.1:

12.1.6 Emulated System OFF mode

If the device is in debug interface mode, System OFF will be emulated to secure that all required resources needed for debugging are available during System OFF, see DIF chapter for more information. Required resources needed for debugging include the following key components: DIF, CLOCK, POWER, NVMC, MPU, CPU, CODE, and RAM. Since the CPU is kept on in emulated System OFF mode, it is recommended to add an infinite loop directly after entering System OFF, to prevent the CPU from executing code that normally should not be executed.

You should always reset the chip after programming if doing power measurements, etc., to make sure it is not in debug mode.

Best regards,
Jørgen