GPIOTE library in SDK7.2 with nRF51422

[Yoshiyuki Kasai]

I'm using GPIOTE library to check the change of two switch inputs. Normally it works well, but rarely GPIOTE_IRQHandler become never happen even if the pin level changes. I think the reason is bounds like noise. For example, the pin changes many times while serving this interrupt.

Could you please tell me the solution?

[Yoshiyuki Kasai]

I'm using SDK library GPIOTE, so I think the event is cleared at the first in the handler. I'm not sure the event is set in GPIOTE, but the input is changed. This is my original circuit. VDD is 2.5V. Two switch inputs is pulled up with VDD. The maximum power noise to VDD is over 7.0V in a few nano seconds. I think this noise make a problem.

[Susheel Nuguru]

7.0V is too far from the operating voltage of 3.6V max, but if the noise is spike for few microseconds then I am guessing that it should not damage the line atleast.

Can you check if the GPIO pin and GPIOTE channel that connected to this pin are operational. You can do this by running GPIOTE example from SDK\examples\peripheral\gpiote and change the code little to match the PIN and GPIOTE channel you used.

If they are operational, then probably the power spike is taking the module into an unknown state. We cannot do much here inside the chip and you need to solve it by external noise reduction circuits outside the pin.

[Yoshiyuki Kasai]

The pin and GPIOTE are operational if the external noise doesn't happen. Thank you for your advice, I try to reduce the external noise.