Potato Cannon Speed Sensor

I built an infrared speed sensor for my air cannon.

240 frames/sec Galaxy S7

I'm not sure if this thing qualifies as a potato gun. It's more like a compressed citrus cannon.

Why?

The cannon is surprisingly powerful, but observing power isn't the same as quantifying it. A couple years ago I tried to quantify this power using a stopwatch...

I shot an orange straight up and recorded its flight time. In a residential neighborhood. Next to parked cars. Turns out straight up is only relatively straight.

Discount artist's rendition

No cars/people/pets were hit, but the fear I realized as the orange began it's descent to earth erased my experimental concentration. Basically I don't remember the results, and didn't iterate.

Download

Arduino code

Video Overview

Lots of video by: John Willner

Design

Break beams! I don't know of a simpler non-invasive method for measuring velocity. Two infrared beams are positioned a set distance apart, and a microprocessor records the time for an object pass between them.

Not broken - logical LOW

Broken - logical HIGH

A voltage divider is formed using the IR detector and a 100k‎Ω resistor. I have added no filtering whatsoever. Hopefully enclosing the beams in a cardboard tube will stop ambient light from affecting the infrared detectors. #NoFilter

First I cut the cardboard tube with a hacksaw, then a laser.

Laser cutting holes for the beam and set screws

Screws are used to mount the cardboard tube to the air cannon.

Sensor mount

Some sortof display to display the readout is needed. I upcycled an old video prop. Here's a post about said prop.

Sacrifice

Having a functional 4 digit 7 segment saved me quite a bit of work. I had to move two of the display wires to different IO pins to free the hardware interrupts.

Arduino Pro Mini

I then wired the IR beams up.

Beam wiring

First the beam was tested, then a strip of tape was wrapped around to block ambient light and hold everything in place.

Testing one beam

Let's write some code!

Programming

Recording the beam break times was easy using two hardware interrupts. The microprocessor then subtracts the two times to find the delta. The velocity can be calculated using the distance between the two beams:

velocity = distance / time

Velocity is displayed in feet per second, which was a little stupid since I kept asking Siri to convert to mph.

The 7 segment display is multiplexed without using resistors for easy wiring. There are more details in that blog post linked above.

Function

Finished!

My #NoFilter didn't pay off; bright ambient light rendered the sensor useless. I applied aluminum foil liberally to block ambient light.

Keeps the government out

The aluminum foil looked horrible, and didn't work at all. Next shot was going to be in the shadiest spot we could find.

Shady

Just kidding.

The underside of this deck will do.

256 fps or 175 mph!!!!!

So.. The sensor works, as long as you're shooting in the shade or at night.

How Powerful?

We know:

Starting position and velocity = 0
end velocity = 78 m/s
end position = 0.95 m
mass 0.069 kg

What we can find:

Average acceleration
Acceleration time
Projectile energy
Average power

Power calculation

Results:

Average acceleration = 328 g's
Projectile energy = 211 joules
Acceleration time = 0.024 seconds
Average power = 9250 Watts ~ 12.4 horsepower

The orange was dead long before hitting the concrete block.

Acceleration and power blew my mind. I would not have guess the cannon is transferring nearly 10kW of power. And 328 g's means the orange is experiencing 222 N or 50 lb of force!

Conclusion

#NoFilter was dumb. I'm going to fix this with a high-pass filter and amplifier.

The cannon produces way more power than I ever thought it would.

I may be filming the cannon with a real actual high speed camera in the near future.

Air Cannon Build Secrets

There's only one thing you need to worry about: how fast can you release the pressurized air?

Sprinkler values are commonly used, but when left stock are a literal bottleneck. You need to carve them out, like removing backflow reduction features. If your cannon is making noises like a dying elephant, it means air is restricted and you're losing power. 90 degree bends don't help either, but a straight valve makes a long cannon.

The End

That's it.

That's all I got.