Foosball, meet LED Sensor

June 15, 2009 1:22 PM

How to rig up a foosball table to measure velocity of goals?

Our company is foosball-obsessed, and I thought a fun side project would be to rig something up inside the goals to measure how fast the ball is moving when it enters the goal. I do mostly software, so hardware/electronics is a fun yet alien world to me. My (uninformed) thought is shining a light on one end of the goal and a light sensor on the other side, and that way measuring how long the light is obscured for. My google-fu is failing me, so any tips on getting started, or resources on where to look next would be appreciated!

posted by mikeyk to Technology (9 answers total)

Another way you could do it is with a force sensor - attach it to a plate that covers the back of the goal. It's a simple conservation of momentum problem to determine the original speed of the ball from the force measurement. The plate should be relatively lightweight.
posted by backseatpilot at 1:30 PM on June 15, 2009

Wouldn't you need to take into account the angle of the shot if you use a force measurement?
posted by Grither at 1:32 PM on June 15, 2009

My senior Physics project was constructing a CCD camera (from nearly scratch) to be used on a telescope. My friend's project was building a laser grid to map the path and velocity of the ball on a foosball table.

He drilled regularly spaced holes down all four sides, put lasers in one side and optical sensors in the other. This fed into a computer that could plot the position and velocity of the ball.

You could get by with a single beam at each goal, which would be at the height of the circumference of the ball (you'll know the ball's width, the time the optical gate is open, and can derive velocity). Because this is dependent on the ball rolling along the table, and won't account for shots that become airborne, you'd get more accurate results with two gates a centimeter or two apart on each goal (measure the time difference between the opening of each gate).

Please let us know how it goes!
posted by spikeleemajortomdickandharryconnickjrmints at 1:32 PM on June 15, 2009

I'd put two lasers and two sensors on the goal gates. Assuming the lasers are positioned at the midlevel of the ball (and most scores aren't airborne) it should be easy to just time the difference between the light breaks and do the math. An arduino board should have the requisite number of digital inputs (4). The lasers can just be dumb lasers.

...stupid lasers...
posted by chairface at 1:54 PM on June 15, 2009

I'm assuming you don't have access to a lot of physics laboratory equipment. I think the absolute simplest way to do this is to drill a hole in the back of the goal, or just remove the ball-catching part of the table, so the ball shoots out the back when a goal is scored. Put a bunch of tape marks on the floor to measure distance. Then the speed will be proportional to the distance it flies out the back:

v = d * sqrt(g/(2*h))

where
v=velocity
d=distance
g=9.8 m/s^2
h=height of goal
posted by Salvor Hardin at 2:28 PM on June 15, 2009

(note - d is the horizontal distance traveled)
posted by Salvor Hardin at 2:28 PM on June 15, 2009

Actually, since you seemed interested in trying some hardware fun, another simple way to do this is the following:

Materials:
Digital oscilloscope
Photodiode (for instance this one).
5V voltage source (battery)
Resistor (~1 kohm)
Small lamp

Connect up the battery, photodiode, and resistor in series. Hook up the oscilloscope probes so they measure the voltage across the resistor. Position the photodiode so the sensor is on one side of the goal. Position the lamp on the other side of the goal so it shines on the sensor.

When the ball goes across the sensor, the oscilloscope will record the change in voltage, and you will easily be able to read the length of time of the peak. Along with the ball's diameter, that'll give you the speed.
posted by Salvor Hardin at 2:39 PM on June 15, 2009

Do you have a budget in mind? Mattel sells at $68 radar gun. You could probably make something for less than the cost of two of these, but it's an option.
posted by exogenous at 2:58 PM on June 15, 2009

Well, velocity kind of sucks as you need to know the distances or time between two points. If you really wanted to do this simple why not just be happy with acceleration? Ft/s^2 or m/s^2 ?

With that you only need a simple cheap force sensor and the equation F=ma, you will know the force and the mass of the balls, thus you get the acceleration....

just a thought...
posted by Black_Umbrella at 5:10 PM on June 15, 2009