search

Freenove / Freenove_Ultimate_Starter_Kit_for_Raspberry_Pi

Apply to FNK0020
Other
561 stars 294 forks source link

Chapter 14 - a suggestion to improve debouncing #9

Closed mikebern closed 6 years ago

mikebern commented 6 years ago

I was looking at the C version of the Relay and Motor code and noticed that if I increase captureTime to 1 sec (long captureTime = 1000), then it becomes impossible to turn off the motor. It's happening because when the motor is running, it is almost impossible for the button to stay in the LOW state for a few hundred milliseconds even if it's being firmly pressed down all the time. It can be easily seen by printing a debug line when reading == HIGH. The schematic diagram and the code are the latest as of Oct 29, 2018.

The problem seems to be similar to one in issue #7 . There is an easy fix though, just adding a capacitor 0.1 uF aka "104" between "+" and "-" of the power supply makes spurious state transitions much less frequent.

So to summarize, the code works as it is now, but adding a capacitor (which is included in the set) makes the system more robust. Just wanted to share this observation.

SuhaylZhao commented 6 years ago

Hi,

Yes, you are right. Adding a 0.1uF capacitance can really make the circuit more robust.

However, the circuit and code has no problem here, and after our tests, even the captureTime is set to 10 seconds, the project still work properly (set captureTime = 10000).

Because releasing the button also requires debouncing. You need waits the same amount of time(captureTime) from releasing to pressing again.

The code running result prompts whether the key is pressed or released, and only when released state is prompted, the key can be pressed again.

In addition, you can try shorting R2, and the test results will be better.

R2 is used to avoid that GPIO output too large current and the IO port is burned when the button is pressed. In this circuit, GPIO18 is input mode, so there is no need for using R2.

Thank you.

mikebern commented 6 years ago

I think you are right. I was actually aware that LOW -> HIGH transition is also debounced, and I have to wait until it comes back to HIGH before pressing the button again. But I forgot to mention about the elephant in the room - I am not running the power source from 9v battery, but from AC/DC wall adaptor, which outputs 9v. I guess it's much noisier than the battery and that makes the critical difference.

So, it looks like I solved the issue that I've created for myself :) Thank you for your prompt follow up and clear explanations! I am closing this issue as it's not a real problem with the circuit.