518822 can't go into sleep mod for current saving

[Songplus]

Hi !

     In order to consume lowest current，we use the function power_manage() ，let the MCU go into sleep，the SD is S130, V1.0，we enable the GPIOTE，watchdog ，ADC ，app_timer，and UART0.Before going into sleep ,we stop(disable)the UART0.We find that sometimes  MCU can go into sleep，but sometimes not. I'm sure that unused IOs are not in float statue. Current sample resister is 5 Ω, the voltage between it is 5~8mV，it meas that current is 1mA.When we debug the program ,find  the internal 16MHz RC osillator is alway running. I *** the power_manage()  does'nt tack affect. How to solutethe problem，our production cann't work in low power mod, it is terrible.There maybe a bug in this chip.

.

[Simonr]

Hi

Please try disabling the GPIOTE peripheral as well, as that will keep the HFCLK running if it's left enabled. You'll have to make sure that all peripherals are disabled before going to sleep mode.

Best regards,

Simon

[Songplus]

Hi，When I disabling the GPIOTE，current consumption upto 4mA，I don't know how to do?

[Songplus]

I test the IC in the smallest system，all GPIO disconnnect.

(1)confige all GPIO pulldown input

nrf_gpio_range_cfg_input(0,31,NRF_GPIO_PIN_PULLUP)

,run t advertising at he S130 . the chip can't work;

(2)Let the GPIO at reset statue,run  advertising at he S130 . the chip can work; current consumption is float ,change from 0.4~1mA.

(3)Confige GPIO as ouput one by one,the chip can work; current consumption is about 1mA, run advertising well;

(4)Confige GPIO as ouput one by one, in the main funtion just do "power_manage();" current consumption is 1mA .There is a plus every 250ms,I don't know what happen in the chip.

The test code as follow:

int main(void)   
{
	volatile static uint32_t err_code;
	bool                     erase_bonds=true;	
	uint8_t                  i;

	for(i=0;i<32;i++)//can work
	{
		nrf_gpio_cfg_output(i);	  //make sure the pin isn't float
	}	
	/*	 
	timers_init();		
   ble_stack_init();
   err_code = pstorage_init();
   APP_ERROR_CHECK(err_code);
	 

 
   device_manager_init(erase_bonds);
   db_discovery_init();
   csc_c_init(); 

   Init_user_info_store();
   load_gps_correct(&m_sys_data_store);	
	 
   ph_device_manager_init(false);	
   gap_params_init();	 
   advertising_init();
   services_init();
   hids_init();
   ancs_service_init();
   conn_params_init();

   Init_CentralDevice_Scan();

   Init_App_Comit();
//   ADC_Config();  //unused  	 
//   nrf_gpio_range_cfg_input(0,31,NRF_GPIO_PIN_PULLUP);//add this function，CPU can't work

   advertising_start();
*/

   
	
   for (;;) 
   {

      power_manage();

		 
	 }

}

So, frome the test ,the chip maybe have a big bug ,it isn't goog for using at low power production .Or my operation maybe exist a mistake.Could Nordic engineers give a good solution? I'm waiting...

[Songplus]

Today ,I look up the SDK revisions，found that SoftDevice S130 V2.0.0 support NRF51822 V3.Now my  production use  S130 V1.0.0,but Chip revision unknow(N51822 QFACA2 1921DC).Is the   S130 V1.0.0 supporte V3? 

[Songplus]

MDK5 Projection in SDK12.3.0 miss too many files.Where  are these files ?  Too many changes in the lastest SDK ,How to  migrate the S130 V1.0.0 to he V2.0.1? Is there a integral project ？If change too much ,we need use a long time to modify our production.