Proximity sensor - metal ball track

I am working on adding some gimmicks to a ball track using metal balls. Looking at the different types of sensors I was thinking that an IR sensor like TCRT5000 might work best, just not sure whether it will accurately detect passing balls. My other idea was an inductive sensor, more expensive but maybe more reliable, less false readings. Anyone's got any experience with a similar project ?

Why not just try?

BBs or cannonballs? Will the balls always be separated by at least their own diameter? How fast are they moving?

I ordered the components and will then start the experiments. Was hoping that some people might have experience with similar projects.

Not sure what BBs or cannonballs are sorry, they are 13mm steel balls. While moving the balls are always separated but there are "waiting areas", for example in front of a magnetic elevator. In those areas, if using more than one ball (which is more fun) they accumulate one after the other, touching each other. In order to measure the speed accurately I would probably first have to have the sensors but I calculated that the speed goes up to about 1.5m/s, will vary a lot obviously.

Your TCRT5000 is a reflective sensor. I would also try a light beam interrupter. At 3/4 of the ball diameter height, so balls that are one after the other can be distinguished.

thanks, great proposal, was actually looking at such initially but I am struggling to find appropriate components, was looking for example at LM393 but it looks like the distance isn't wide enough for the balls (only 10mm)... might still work if just measuring upper part of balls. Do you know of other components that might be a better fit ?

Just use a separate emitter and detector.

If you choose an optical sensing method it often helps to offset the sensor angle a bit from perpendicular to the line of travel of the objects.


Also commonly used: microswitches with levers at DuckDuckGo

thanks !

ok, I will check the microswitches as well... thanks !

do you have a proposal for the components, struggling to find, sorry I am pretty new to electronics... I might check if white LED and photosensor on each side of the track might work... bit of a crude solution.

What is crude about that? I wouldn't recommend a white LED though... those are inefficient due to the phosphor conversion, and also doesn't produce much energy in the longer wavelengths, like red and infrared, where photodiodes are usually most sensitive.

My garage door has an optical beam sensor, works very well over 3 meter range.

Metal balls? Perhaps you should look into a Hall Effect sensor.

With a bias magnet.

I found a Hall effect sensor ineffective as well as a magnetic field sensor.

The three methods that worked was a interrupt beam of various sorts, a light sensor detecting the ball as it rolled over and a time of flight distance sensor.
Video here
Gravitrax aproject

See where to get a free download of the article published in the Mag Pi magazine. Also look out for part two.

Thanks Mike, nice project ! Would you mind giving me the specs of the sensors you used ? What exactly is a time of flight distance sensor ?

The full range of sensors I used is in the magazines.
Issue 87
Issue 88
Issue 89

It sends out a pulse of IR light and times how long it takes to reflect off an object and return. Google "TOF sensor".
There are a few types. This is one of them.

cool, thanks for the links... the TOF sensor looks really cool, not cheap though but I will give it a try.