Pressing button fast leads to 20uA current consumption in system_on sleep mode.

To debug the problem, I have tried to remove all codes in button_event_handler, so that nothing could influence the APP_button library, and I got the same result: after pressing the button very fast, then CPU gets into SYSTEM_ON sleep, the current consumption is 20uA.

Hi,

1) This is not the peak current, but a state-state 20µA current consumption, correct ?

2) After it got into this state, does it go back to 2uA if you press the button again?

1) Yes, it is a steady state 20uA current consumption during SYSTEM ON sleep. It can stay in 20uA the whole day, it won't get back to 2uA by itself.

2) When it is 20uA, I tried to keep fast/slow pressing the button (to wake up the chip and put it into sleep once again), I found sometimes it can get back to 2uA, but very rare. 

I don't have a Development Kit by my side, so I can't test this on a DK.

Hi, I think I have found the problem, but I need the guide of how to fix it.

What I have found is, the CPU gets into SYSTEM ON sleep while the EVENTS_COMPARE[0] register of RTC1 (app_button's timer) is set. (I have a screen shot of KEIL register viewer)

Isn't it possible that app_button library doesn't clear the EVENTS_COMPARE[0] register under this case?

And I wonder if there is a proper way to clear it? I have read the datasheet, the RTC's TASK_CLEAR only clear the RTC counter register.

Thank you.

Hi, I think I have found the problem, but I need the guide of how to fix it.

What I have found is, the CPU gets into SYSTEM ON sleep while the EVENTS_COMPARE[0] register of RTC1 (app_button's timer) is set. (I have a screen shot of KEIL register viewer)

Isn't it possible that app_button library doesn't clear the EVENTS_COMPARE[0] register under this case?

And I wonder if there is a proper way to clear it? I have read the datasheet, the RTC's TASK_CLEAR only clear the RTC counter register.

Thank you.

This sounds strange.

1) Are you using app timer v2? If not, try using app timer v2

2) Could you upload a small example that triggers this issue?

Correct, the KEIL has APP_TIMER_V2 define.

I'd like to upload screenshots and example project, but it needs a external storage space link?

Another question is: how could I clear EVENTS_COMPARE[0] register in my code? Thank you.

gilbertjuly said:

I'd like to upload screenshots and example project, but it needs a external storage space link?

 
Zip the project, and upload it like this:

gilbertjuly said:

Another question is: how could I clear EVENTS_COMPARE[0] register in my code? Thank you.

 The driver already does this. In app_timer2.c -> rtc_irq ->drv_rtc_compare_pending -> evt_pending -> nrf_rtc_event_clear()

hi, attached is the example project:

The test environment is SDK 17.02.

The test case is:

1. Test the system-on sleep current consumption. 

2. The code has three buttons, press any button very fast, like 3 times per second.

3. After the CPU gets into system-on sleep again, test sleep current consumption.

In step 3, we can find it has 20uA additional current consumption compared with step 1.

I don't have a DK, regarding the key circuit, my board is designed like this:

I attempted to debug, what I have found is: the CPU gets into SYSTEM ON sleep while the EVENTS_COMPARE[0] register of RTC1 (app_button's timer) is set. Attached is the KEIL register viewer during code running at "wait for event".

I hope you can reproduce this. Thank you in advance.

Jiayan

Hi,

gilbertjuly said:

I hope you can reproduce this. Thank you in advance.

Update: I reproduced it on a DK. Looks like m_pin_active in app_button.c is stuck high, and the m_detection_delay_timer keeps getting started each time it times-out. I have not found a fix yet.