Power consumption of NRF52 GPIOTE->IN

RE: Power consumption of NRF52 GPIOTE->IN

gilbertjuly — Tue, 08 Jun 2021 09:07:44 GMT

Thank you so much.

RE: Power consumption of NRF52 GPIOTE->IN

ovrebekk — Mon, 07 Jun 2021 09:49:56 GMT

No, the difference between one and two GPIOTE IN events should be minimal.

Whether or not you add one or more GPIOTE IN events the current consumption will increase by approximately 20uA.

Best regards
Torbjørn

RE: Power consumption of NRF52 GPIOTE->IN

gilbertjuly — Fri, 04 Jun 2021 15:52:11 GMT

Hi Torbjørn, thank you for reply.

One GPIO's GPIOTE IN event increases the current by 20uA...

Does it mean: two GPIO's GPIOTE IN events increase the current by 40uA? And three increasing by 60uA?

RE: Power consumption of NRF52 GPIOTE->IN

ovrebekk — Fri, 04 Jun 2021 11:55:57 GMT

A GPIOTE IN event on the nRF52 should increase the sleep current by around 20uA.

In general though please open a new ticket if you are having issues with this, since this case is quite old.

Best regards
Torbjørn

RE: Power consumption of NRF52 GPIOTE->IN

gilbertjuly — Thu, 03 Jun 2021 03:19:06 GMT

Then, how much current draw will nRF52 cause when using GPIOTE IN event (high precision)?

RE: Power consumption of NRF52 GPIOTE->IN

Ole Bauck — Tue, 13 Sep 2016 08:57:26 GMT

See this post. If you are using SPIM or TWIM in combination with GPIOTE->IN you may experience an increase in current of about 450uA. This is related to this errata (it does not mention SPIM since we are working on figuring out what the underlying cause is).

RE: Power consumption of NRF52 GPIOTE->IN

nrf52 — Tue, 13 Sep 2016 06:25:42 GMT

Hi Ole, As you said "With GPIOTE_CONFIG_IN_SENSE_HITOLO(true), which uses GPIOTE->IN you may see an increase in current consumption of ~450uA" , I meet the same issue, when enable GPIOTE->IN interrupt , increased "~450uA" on nRF52 board, how to fix it ,thanks

RE: Power consumption of NRF52 GPIOTE->IN

Jorgewu — Wed, 22 Jun 2016 06:51:25 GMT

After check my codes, it seems follow design requirements. So, I am asking HW engineer to breakdown power consumption of each component..

RE: Power consumption of NRF52 GPIOTE->IN

Ole Bauck — Tue, 21 Jun 2016 10:13:52 GMT

If you use IN event (high_acc = true) on nRF51 you will get around 800-900 uA. This is because the IN event on nRF51 request the HFCLK and 1V2 regulator as seen in the Product Specification part 8.3. If you use PORT event the current draw should be reduced to almost zero. This is done by setting it to false.

On nRF52 this is different. IN event on nRF52 should not result in higher current draw.

RE: Power consumption of NRF52 GPIOTE->IN

soumil — Tue, 21 Jun 2016 06:23:11 GMT

Just adding to the discussion here, even I am facing similar issues of high sleep current. I am using nrf51882 S110 with SDK 8. On using the GPIOTE_CONFIG_IN_SENSE_TOGGLE(high) function i am getting a sleep current of 860 micro amperes. Is it because this function doesn't use the PORT event and if i set to false will it reduce the sleep current?

RE: Power consumption of NRF52 GPIOTE->IN

Ole Bauck — Mon, 20 Jun 2016 11:43:07 GMT

You are sure the 1.9mA is not caused by other parts (e.g. chips) in the system? The nRF52 should not be able to draw 1.9mA in system OFF.

RE: Power consumption of NRF52 GPIOTE->IN

Jorgewu — Sat, 18 Jun 2016 12:18:45 GMT

I implemented the function to vibrate when system is reset , and wake up system by Gsensor interrupt. After I call the function of power off after broadcast timeout, there is no vibration happens and it only vibrate after I shake device to trigger G-Sensor interrupt . According this symptom, it's seems system had enter power off mode already and waked up by interrupt. it it true? However, the power consumption still keep 1.9mA.. Why ???

RE: Power consumption of NRF52 GPIOTE->IN

Ole Bauck — Fri, 17 Jun 2016 14:24:07 GMT

Try to turn off UART and see if the current consumption is less. You are sure that the sensors are not drawing current? The reset reason have to be cleared manually, it is not cleared unless the reset is power on or brown out reset, see here. Your reset reason shows many things (pin reset, soft reset and wake up from system off reset).

RE: Power consumption of NRF52 GPIOTE->IN

Jorgewu — Fri, 17 Jun 2016 02:15:39 GMT

After fine tuning my codes the power is down to 1.97 mA in the conditions(The current is Same as the conditions of just execute the loop of __WFI(), in the main()) 1.Enable softdevice. 2. without advertising.. 3. My HW just connect G-sensors via I2C However it still cannot enter power off mode... My Sense pin is pin27 , here is the pin status before enter power_off and reset reason(UART is enabled to print message(pin30, 31 for CTS/RTS, and pin 29 for DTM enable )) NRF_GPIO->IN:0x80200000 Reset Reason:0x50005 Is any item needed be checked again?

Thanks..

RE: Power consumption of NRF52 GPIOTE->IN

Ole Bauck — Thu, 16 Jun 2016 15:09:32 GMT

If you use GPIOTE_CONFIG_IN_SENSE_HITOLO(false) you should not see any increase in current, this uses GPIOTE->PORT not GPIOTE->IN. With GPIOTE_CONFIG_IN_SENSE_HITOLO(true), which uses GPIOTE->IN you may see an increase in current consumption of ~450uA.

1.9mA when not doing anything is very high. Are you advertising or scanning often or doing other power hungry stuff?

If you enter system OFF the current consumption should be less than 1uA. If you see the same current consumption after going to System OFF the device probably have woken up from system OFF mode. If the pin configured for GPIOTE_CONFIG_IN_SENSE_HITOLO(false) is low when you enter system OFF the chip will immidiately wake up (GPIOTE->PORT can wake the chip up from system OFF sleep).

RE: Power consumption of NRF52 GPIOTE->IN

Jorgewu — Thu, 16 Jun 2016 14:09:14 GMT

If I enabled the GPIOTE->IN the current is ~3.1mA(2.8V) by const nrf_drv_gpiote_in_config_t in_config1 = GPIOTE_CONFIG_IN_SENSE_HITOLO(true). And the current decreased to ~2.6mA after set the config by GPIOTE_CONFIG_IN_SENSE_HITOLO(false). If I have not enable GPIOTE->IN the current is ~1.9mA..

BTW,

Is the current consumption of GPIOTE->IN still keep 500uA while enter power off after enable it? if I just execute for loop to call power_manage() in the main(), the current is 1.9mA, is it reasonable? 3.To reduce power cinsumption, I tried to enter power off mode by call the function of sd_power_system_off(). However, the power consumption always same as before execute the function.. Is any item needed be prepared before enter power off mode?

Thanks..

RE: Power consumption of NRF52 GPIOTE->IN

Ole Bauck — Thu, 16 Jun 2016 11:01:57 GMT

We have seen problems when using GPIOTE->IN combined with other peripherals that resulted in a current increase of about 450uA due to the high frequency clock being requested, but not 1mA. Can you give us some more information about your application?