GPIO signal: expect "high" as magnet detection

[ghost]

Asking for help locating the part of the GPIO code where the detection of the magnet would be reversed. Why? I wanted to go away from mechanical hall sensors and try something like TL4905L or a A3144 board with LM393 comparator that gives the"high" state when magnet is detected. Such board is even cheaper than some of hall effect sensors and more reliable and some of them even have a regulator (potentiometer) where you can precisely adjust the sensitivity of the circuit.

In ideal case this setting could be as one of the parameters in the config file too.

[laberning]

That's an interesting idea! Have a look at openrowingmonitor/app/gpio/GpioTimerService.js to change the gpio sensor settings.

All that you would have to change is the initialization call for the sensor:

// read the sensor data from one of the Gpio pins of Raspberry Pi
    const sensor = new Gpio(config.gpioPin, 'in', 'rising')

For your scenario you probably want to set it to falling. There are also a bunch of additional options (like denouncing) that you could set if needed, see here for the available options: https://github.com/fivdi/onoff#gpiogpio-direction--edge--options

If it works out nicely we can make it a configurable option and describe the sensor setup in the docs...

[ghost]

Wow! Thanks. I was watching yesterday night at this part, but wasn't sure if the 'falling' is all it needs. I am not JS developer (more like Linux console enjoyer or Arduino fan and such).

I bought recently a reed sensor but it seems, that this part is very delicate and I've managed to broke three of them and finally gave up because I didn't had more. After that spectacular failure, I've ordered a simple Arduino module that has A3144 with LM393 and also another one optical reflective sensor with IR LED and IR receiver with adjustable sensitivity, but both give "high" when magnetic field is detected / light detected.

The idea behind this cheap $1 optical reflective sensor is that the flywheel in my case can easily be equipped either with printed out white/black stripes or sticked with reflective sticker for that (even easier). The precision behind detection of black / white is more than detecting magnetic field and more to that - we can decide the length of the impulse by modifying the length of the stripe. Even more to that: we can even put 100 stripes and precisely count them all and have very precise data. I know, it's overkill and we would operate within milliseconds between impulses during heavy workout, but I wanted just hint, that even that is possible and why this way would be even better than mechanical Hall effect sensor.

As I am waiting during the weekend for my new parts, I would post in another topic my modification of the rower with pictures and details.

[JaapvanEkris]

In the proposed modifications, the option to select an upward, downward or both flanks has been added (see #84).