Over-Thought Potato Gun
Over-Thought Potato Gun
Embedded Files
Started 11/1/20
Started 11/1/20
As with most of my other projects, I was surfing YouTube one day when I saw a slow-motion video of a potato gun being fired. I thought to myself, "Hey, potato guns are pretty simple and I'm sure it would be fun to make a really high-quality one." I also saw an article on the MIT technology review which summed up a study done by the Air Force on the internal ballistics of a potato gun. Why exactly the Air Force studied the ballistics of a potato gun? I do not know. Nevertheless, it's pretty funny.
As with most of my other projects, I was surfing YouTube one day when I saw a slow-motion video of a potato gun being fired. I thought to myself, "Hey, potato guns are pretty simple and I'm sure it would be fun to make a really high-quality one." I also saw an article on the MIT technology review which summed up a study done by the Air Force on the internal ballistics of a potato gun. Why exactly the Air Force studied the ballistics of a potato gun? I do not know. Nevertheless, it's pretty funny.
I started drawing something and headed to the nearest hardware store.
I started drawing something and headed to the nearest hardware store.
There are two main types of potato guns: compressed air and combustion-driven. Compressed air guns are, in my opinion, very dangerous and much more expensive than the latter option. The idea behind them is pretty simple: they have a high pressure air reservoir connected to a barrel with a potato in it. The main caveat is the fact that the air inside the reservoir is essentially trying as hard as it can to get out. That means that every piece is a link in a chain, and if one breaks, you're going to get shrapnel. For this reason, I went with a combustion-driven one (granted, when you have combustion in a pipe, you create a pipe bomb if you're not careful). Most of the ones out there use a 4" diameter PVC tube as a combustion chamber with a 2" diameter barrel. By spraying hairspray (or another aerosol) into the chamber, sealing it and then igniting it some way, "thing go boom." I don't really like this idea, so here is my journey detailing what I did.
There are two main types of potato guns: compressed air and combustion-driven. Compressed air guns are, in my opinion, very dangerous and much more expensive than the latter option. The idea behind them is pretty simple: they have a high pressure air reservoir connected to a barrel with a potato in it. The main caveat is the fact that the air inside the reservoir is essentially trying as hard as it can to get out. That means that every piece is a link in a chain, and if one breaks, you're going to get shrapnel. For this reason, I went with a combustion-driven one (granted, when you have combustion in a pipe, you create a pipe bomb if you're not careful). Most of the ones out there use a 4" diameter PVC tube as a combustion chamber with a 2" diameter barrel. By spraying hairspray (or another aerosol) into the chamber, sealing it and then igniting it some way, "thing go boom." I don't really like this idea, so here is my journey detailing what I did.
This is my first drawing of the system. The plan was to measure out specific amounts of propane from a blow torch and deliver it into the combustion chamber. I also toyed with the idea of adding extra oxygen as well, but I realized that starts to get into the territory of a potato howitzer. It is, however, still up in the air (pun intended) for future experimentation. I got some brass fittings and got to work.
This is my first drawing of the system. The plan was to measure out specific amounts of propane from a blow torch and deliver it into the combustion chamber. I also toyed with the idea of adding extra oxygen as well, but I realized that starts to get into the territory of a potato howitzer. It is, however, still up in the air (pun intended) for future experimentation. I got some brass fittings and got to work.
By knowing the volume of the container, adding propane to a specific pressure can yield the Goldilocks stoichiometric ratio between fuel and air.
By knowing the volume of the container, adding propane to a specific pressure can yield the Goldilocks stoichiometric ratio between fuel and air.
Here's the first setup I used. The propane tank attaches to the torch on the top right, which injects fuel into the tube up to the ball valve on the left side of the image. I initially attached the ball valve directly to the 4" screw cap, but it got really annoying to deal with when screwing it on to the combustion chamber. I would disconnect the propane from the air compressor couplings in the center of the photo.
Here's the first setup I used. The propane tank attaches to the torch on the top right, which injects fuel into the tube up to the ball valve on the left side of the image. I initially attached the ball valve directly to the 4" screw cap, but it got really annoying to deal with when screwing it on to the combustion chamber. I would disconnect the propane from the air compressor couplings in the center of the photo.
In the new version below, I attached a female compressor fitting to the cap, allowing not only a quick release, but the ability to turn the rest of the brass freely when attached. It also allows for much easier positioning of the cannon when launching.
In the new version below, I attached a female compressor fitting to the cap, allowing not only a quick release, but the ability to turn the rest of the brass freely when attached. It also allows for much easier positioning of the cannon when launching.
With the fuel-injection system ready, I went on to making an ignition system. At first, I used an ignition coil I had that was waiting to be implemented in a project, powered by a 6V battery(left). By rapidly connecting and disconnecting the battery, the secondary coil produces a few kV which can arc across a spark gap (right).
With the fuel-injection system ready, I went on to making an ignition system. At first, I used an ignition coil I had that was waiting to be implemented in a project, powered by a 6V battery(left). By rapidly connecting and disconnecting the battery, the secondary coil produces a few kV which can arc across a spark gap (right).
Rather than using a relay, I 3D printed a thing (for lack of a better word) on which I can drag an alligator clip or other piece of wire to rapidly connect and disconnect the battery, producing sparks. For my very first test, I brought the launcher onto my patio and aimed at some bricks. I filled the delivery system to the pressure that I had calculated it needed to be, took cover behind a chair and started sparking the coil. Nothing happened, and I figured that was due to the fact that I overlooked what the purity of the propane I had may have been. The next logical step was to "pump more gas 'till she blows," and that's what I did. On the next try, there were pieces of potato all over the place.
Rather than using a relay, I 3D printed a thing (for lack of a better word) on which I can drag an alligator clip or other piece of wire to rapidly connect and disconnect the battery, producing sparks. For my very first test, I brought the launcher onto my patio and aimed at some bricks. I filled the delivery system to the pressure that I had calculated it needed to be, took cover behind a chair and started sparking the coil. Nothing happened, and I figured that was due to the fact that I overlooked what the purity of the propane I had may have been. The next logical step was to "pump more gas 'till she blows," and that's what I did. On the next try, there were pieces of potato all over the place.
My mom mentioned to me that she'd find it safer if I could remotely fire it, so I obliged. My plan was pretty straightforward: I broke out two Arduino Nanos I had as well as two transceiver modules. Using a MOSFET, I would be able to switch the coil on and off digitally to send sparks.
My mom mentioned to me that she'd find it safer if I could remotely fire it, so I obliged. My plan was pretty straightforward: I broke out two Arduino Nanos I had as well as two transceiver modules. Using a MOSFET, I would be able to switch the coil on and off digitally to send sparks.
On the left is the remote, and on the right is the barrel of the spud cannon. In the middle is a muzzle velocity "measureinator," which I'll get into in a moment. The code is straightforward: If a button is pressed on the remote, a signal tells the receiver to turn on the FET.
On the left is the remote, and on the right is the barrel of the spud cannon. In the middle is a muzzle velocity "measureinator," which I'll get into in a moment. The code is straightforward: If a button is pressed on the remote, a signal tells the receiver to turn on the FET.
The ignition coil proved to be unreliable, so I turned to good old 2,500*F melting point nichrome wire. It worked, but at the same time it sucked, due to the amount of time it took to warm up to 300*C (propane autoignition temperature). I ended up buying a super high voltage boost converter which sparks reliably on command. These things really shouldn't be available to buy on Amazon.
The ignition coil proved to be unreliable, so I turned to good old 2,500*F melting point nichrome wire. It worked, but at the same time it sucked, due to the amount of time it took to warm up to 300*C (propane autoignition temperature). I ended up buying a super high voltage boost converter which sparks reliably on command. These things really shouldn't be available to buy on Amazon.
With an ATmega strapped to the launcher, I thought it would be reasonable to make something which could measure the muzzle velocity of a potato. My first instinct: lasers. I ordered photoresistors and laser diodes from Amazon and got to work. I designed and 3D printed a muzzle to hold everything together.
With an ATmega strapped to the launcher, I thought it would be reasonable to make something which could measure the muzzle velocity of a potato. My first instinct: lasers. I ordered photoresistors and laser diodes from Amazon and got to work. I designed and 3D printed a muzzle to hold everything together.
This is the first muzzle I made. There were two laser diodes shining onto two separate photoresistors. When a potato passes through a beam, a resistor would change resistance which could be picked up by a comparator. By measuring the time it takes for a potato to travel between the beams, a velocity could be calculated.
This is the first muzzle I made. There were two laser diodes shining onto two separate photoresistors. When a potato passes through a beam, a resistor would change resistance which could be picked up by a comparator. By measuring the time it takes for a potato to travel between the beams, a velocity could be calculated.
It turns out that photoresistors are VERY slow, with a 50-100ms response time. In order to be accurate, I needed response times down to a handful of microseconds. Before even testing the laser/resistor setup, I went back to the drawing board. I did some research and found out that LEDs also function as photodiodes, producing a small current when the correct wavelength is shined onto them. I got four LEDs I had left over from the LED cube and got to work.
It turns out that photoresistors are VERY slow, with a 50-100ms response time. In order to be accurate, I needed response times down to a handful of microseconds. Before even testing the laser/resistor setup, I went back to the drawing board. I did some research and found out that LEDs also function as photodiodes, producing a small current when the correct wavelength is shined onto them. I got four LEDs I had left over from the LED cube and got to work.
I remade the muzzle to be slightly larger, increasing the distance between the LEDs to 60mm. Increasing the distance increases the time the potato takes to cross both beams, lowering the error due to imperfect timing.
I remade the muzzle to be slightly larger, increasing the distance between the LEDs to 60mm. Increasing the distance increases the time the potato takes to cross both beams, lowering the error due to imperfect timing.
As you can see in the photo above, the muzzle is wired to a comparator (an LM393 with a 1.3us response time. 2 comparator circuits in 1 IC), (future me here, I should've used a transimpedance amp) which I set to go LOW if a beam was interrupted. This LOW signal was tied to an interrupt on the Arduino Nano, which takes note of the current time since the program began. Once the other beam gets interrupted, another interrupt on the Nano takes note of the time again, finds the difference between that and the first time, calculates the velocity, and reports it back to a computer. Here are some of my tests:
As you can see in the photo above, the muzzle is wired to a comparator (an LM393 with a 1.3us response time. 2 comparator circuits in 1 IC), (future me here, I should've used a transimpedance amp) which I set to go LOW if a beam was interrupted. This LOW signal was tied to an interrupt on the Arduino Nano, which takes note of the current time since the program began. Once the other beam gets interrupted, another interrupt on the Nano takes note of the time again, finds the difference between that and the first time, calculates the velocity, and reports it back to a computer. Here are some of my tests:
filtered-7D07C7EF-8F37-43B0-9F99-AA0057AE972E.mp4IMG_4373.movIn case you missed that shot above, here's a freeze frame:
In case you missed that shot above, here's a freeze frame:
For one of my trials, I measured a muzzle velocity of 40m/s (89.5mph), and then again with slightly more propane at 50m/s (111.8mph). The results given by the study indicate that acetylene blows the other propellants out of the water. Acetylene is much more dangerous to deal with than propane, but I would love to test it out in the future.
For one of my trials, I measured a muzzle velocity of 40m/s (89.5mph), and then again with slightly more propane at 50m/s (111.8mph). The results given by the study indicate that acetylene blows the other propellants out of the water. Acetylene is much more dangerous to deal with than propane, but I would love to test it out in the future.
A note about accuracy: To make sure that my velocity measurer was accurate, I tested it by dropping a golf ball through a piece of PVC into the muzzle. Knowing the height at which I dropped the ball allowed me to calculate the velocity that the ball should be traveling at that point thanks to kinematic equations. My calculations and measured velocities lined up, so I am confident in the muzzle's accuracy. That being said, I have conducted too few trials to reach a conclusive result, meaning the 50m/s is, at best, an estimate.
A note about accuracy: To make sure that my velocity measurer was accurate, I tested it by dropping a golf ball through a piece of PVC into the muzzle. Knowing the height at which I dropped the ball allowed me to calculate the velocity that the ball should be traveling at that point thanks to kinematic equations. My calculations and measured velocities lined up, so I am confident in the muzzle's accuracy. That being said, I have conducted too few trials to reach a conclusive result, meaning the 50m/s is, at best, an estimate.
Page updated
Google Sites
Report abuse