Interrupt edge minimum speed requirements?

[valentin]

Posted on April 13, 2016 at 01:28

Hi,

I'm currently thinking about anti-bounce for button inputs on my design and this brought me to the above question.

As I'd like to avoid adding an extra schmitt-trigger for cost reasons, I'm doing the anti-bounce using a simple RC low pass filter.

Does anybody know what the minimum rising (or falling) speed of the edge must be for it to be detected by the NVIC?

http://tronixstuff.com/wp-content/uploads/2010/06/rccharge.jpg?w=300

[Walid FTITI_O]

Posted on April 14, 2016 at 12:40

Hi Valentin,

It is mentioned In the datasheet of the device you are using (see I/O AC characteristics table)

For example for STM32F4xx devices,  The minimum Pulse width of external signals detected by the EXTI controller (

tEXTIpw

) is 10 ns.

-Hannibal-

[valentin]

Posted on April 18, 2016 at 01:57

Hi Hannibal,

thanks for the hint. I had overlooked that part.

I take it all inputs have an internal schmitt-triger like the reset pin?

As my concern was actually about the rising edge being too slow to be detected. It's working fine during testing but I'd simply like to know if there's any upper limit on rise time so I can stay at a safe distance to that.