I'm thinking of building an Arduino device for measuring pitching speed to add some legitimacy to our backyard baseball games. I'm having some trouble coming up with a suitable mechanism so I thought I'd see if anyone here had any good ideas.
I was thinking of using a couple of lasers and phototransistors to set up a gun chronograph sort of system. I've made a proof of concept for this using a length of 1" PVC and a metal bearing. The one problem I'm running into is that neither me nor any of my buddies are major league pitchers, so our accuracy is not so hot. This means we'd probably end up throwing the ball a few dozen times before actually getting it to pass through the required space to break both beams and take a measurement.
So, does anyone have any idea for how I could expand the "measurable area" for the chronograph? Obviously I'm not expecting to be able to hit the target everytime, but being able to get a reading once every 3 or 4 throws would be nice.
Firearm chronometers use an upward-pointing light sensor that detects the tiny change in light intensity when the projectile flies between the sensor and the sky. Perhaps a photo-transistor and a cylindrical lens could produce such a sensor curtain. A high-pass filter would distinguish a passing ball from slow changes in ambient lighting. Place two curtains a known distance apart and time the flight across that distance.
Perhaps the two interrupt inputs on the Arduino would be good for this. Grab the current time in microseconds as each curtain triggers. If you get triggers in the right order you can easily calculate the elapsed time.
I'd install two lasers (output line-form beam), known distance apart, two lines in parallel each other pointing to sky. Next, in the middle would be a video camera also pointed to the sky . When ball cross two parallel flat surface, camera would register two flashes, visible as dots on the image, a several frames apart. Speed could be calculated precisely, as trajectory of the ball would be known in 3D, distance = sqrt(X^2 +Y^2 +Z^2);
time = Numbers of frame x period; V= D/T.
There is a project, that shows and idea of distance measurements using laser: http://fftarduino.blogspot.ca/2011/12/arduino-laser-3d-tracking-range-finder.html
I would think using a range sensor would be your best bet. Something like the S404 (I think it is) ultrasonic generates a cone, rather than a beam. A couple sensors going simultaneously, each taking readings every few milliseconds, should give you the numbers needed to do the math. There's ultrasonic rangefinders and ir rangefinders on the market, all which sit in the $10/pair I believe.
The firearms chronograph approach gives you a space of almost a foot where the projectile has to run, using the proper lenses.
But since baseball pitches are relatively slow you might try ultrasound.
Firearm chronometers use an upward-pointing light sensor that detects the tiny change in light intensity when the projectile flies between the sensor and the sky. Perhaps a photo-transistor and a cylindrical lens could produce such a sensor curtain
John, exactly. Making the "curtain" is one of the problems I'm having conceptually. I just had a look at how gun chronographs work after reading your suggestion and I do think that your idea is a better way to go than what I've come up with at this point (having a laser shining at a spinning mirror, with a strip of 20 photodiodes directly opposite). I'll see what I can come up with as far as sourcing lenses to widen the collection area for the transistor.
dirtyharry2, I had considered ultrasonic, but how accurate are those typically? I'd imagine the price rises with increasing accuracy. I like the idea of a gun chronograph type of setup simply because getting fairly accurate results is pretty simple conceptually as opposed to a rangefinder where you're are the mercy of the rangefinder device.
magician, Interesting. I hadn't even considered anything besides the horizontal velocity. That is a very good point though, and might be something to incorporate in the future if I ever get this thing running.
I forgot one important requirement: the device should probably NOT break if the pitch isn't on the mark ;). Thanks for all the input guys, I'm going to think about what you all have said and try to come up with some idea. Please feel free to chime in if you've got any more ideas.
Does it need to be a contactless measuring system?
How about getting a sheet of some flimsy material that the ball can pass through, and have a sensor to detect when that happens? Perhaps an aluminium foil panel with some way to measure the resistance of a few paths through it? With two of these a short distance apart, you'd be able to do a simple distance/time calc to get the speed.