Project Hovercraft

The goal of this project was to build a working remote controlled hovercraft while spending as little money as possible. I was successful, as the hovercraft works (mostly) and the only money I spent was on the batteries and propellers.



The workings of a hovercraft are fairly straightforward: one high-power motor with an airplane propeller forces air down through the hole in the center, which pushes the bottom of the hovercraft off the ground. This greatly reduces friction, allowing the hovercraft to scoot around without much trouble. It also makes it much more difficult to control, however: when the hovercraft turns, it will keep traveling in a straight line unless more thrust is applied. Since we are used to controlling cars or boats, the newtonian behavior of a hovercraft is challenging indeed.



The body of the hovercraft is made out of styrofoam, cut with a saw and sanded to smooth the edges. The skirt on the bottom is made out of pieces of a plastic GAP bag, attached with Scotch tape and hot glue. The radio control unit was scavenged from an old boat I had, along with the drive motor mounted on the wooden supports, while the lifting motor was from my physics teacher. The battery is 300 mAh 6V NiMH (or NiCd, I forget), and I have a pair of them. They were about $6-$10 each (again, forgot) and about $12 for the 2 1/2 hour charger. The servomotor for steering was from an old RC car I had, but I unfortunately wasn't able to find the radio unit or else I would've used that instead. I needed the gears in the thing so it could provide enough torque, however.
Probably the single most useful thing during construction was hot glue - the entire thing is held together with the stuff and it will hold for some time to come. The physical construction of the hovercraft was really the easy part... the hard part was trying to get it to work the way I wanted it to. I was plagued with numerous problems, most of them stemming from the cheap RC unit I was using.



The first problem was with the steering. The boat had used a pair of magnets and a pair of electromagnets to steer, flipping a little rudder back and forth by toggling those. The immediate problem I saw, before I even took the boat apart, was that it would only steer when the other (drive) channel was active. This isn't the worst of problems, but it is annoying. The problem I found with the steering method was that it used two signals to control two electromagnets, and I wanted to run one motor in two directions instead. To do this, I had to build the interface board sitting on the front left of the hovercraft, consisting of two transistors. I had originally wanted to use an ungeared motor, so the rudder would move to either side and return to the center from the force of the air being blown backwards. An ungeared motor didn't provide enough torque, however, and it was a bad idea to hold a motor stalled while the craft was turning. So, I took the servo from an old Street Comber RC truck I had and used that for steering. The problem now, however, is that it doesn't center by itself. You can see four wires coming out of the side of the servo, that are not being used. Those four wires are feedback wires that can (very roughly) give me the position of the rudder. However, I am way way way too lazy to design and build a more complicated board that would center it for me (that, and I determined that it wouldn't be able to center it accurately at all, so it would be a waste).
Thus, for now, the steering works. However, because it doesn't automatically re-center the rudder when you let go of the stick, the hovercraft has a major tendency to spin out of control, especially considering how hard it is to pilot already.



The next problem was with the drive motor. I want it to be able to go forwards and backwards, since without this ability the hovercraft gets stuck fairly easily. In this regard, I was not successful - it only goes forwards. The boat was originally run on 9V, so the way they made the motor run in two directions was by connecting one terminal to 3V and alternating the other between ground and 9V. The motor would run fast in one direction and slow in the other. Since the hovercraft runs off of 6V, this isn't really a possiblity. I tried doing the same thing, connecting one terminal to 2.4V (the battery is 5x1.2V cells), but it ran far too slowly in both directions. I then tried building an interface board like I did with the steering, and it definitely should have worked. I fried standard transistors the moment I turned it on, so something was wrong there. I figured there may have been too much current, so I tried power transistors. Before those fried as well, it still didn't work, so I figured something is just weird with the way the signal out works from the RC board. So, I just connected one terminal of the motor to ground and the other to the signal - when you push forward on the stick, it runs at the full 6V, and when you pull back, nothing happens.



Skirt with hovercraft on (left) and off (right)

When I said that none of the problems had been with the physical construction, I lied. One of them was - the skirt. The skirt is the bag that fills up with air, so that the bottom of the hovercraft can deform to match whatever surface it is driving over. Now, it works great, but it was not always thus. My first attempts at a skirt were with a flawed design - I figured I could just fold it in and tuck it under the hovercraft. This repeatedly failed, and it was not until I tried some other things that I had any luck. Turns out the secret is that I actually had to pull the skirt back up underneath the hovercraft, to prevent it from billowing out to the sides. To this end, tape and hot glue were very useful. If you were to look at a cross section of the skirt when powered, it would look like a J (whereas my first attempts resembled more of a backwards L). Having driven it outside, it has gotten a little banged up and is no longer working as well as it once did. However, the hovercraft can still drive over a variety of surfaces and go over obstacles around 1.5 cm in height. I have yet to try it on water, but I hear that the friction slows it down significantly.


With the skirt problem resolved, the drive problem half resolved, and the steering still quite impossible, the hovercraft is done. I'm satisfied with the result, it tought me a lot and stuff, and it's cool (the latter being the main criteria for anything I build). Maybe someday I'll build a bigger one - I have a 1 HP cordless drill motor sitting around that's just begging to be put to good use. We shall see.








Closeup of the front circuitry. The entire RC board is coated in wax (since it was from a boat), and that saved me from frying it a few times... lucky me. You can also see the capacitors I added to the lift motor for noise reduction. Before I added those, whenever the lift motor was turned on it would make a bunch of RF noise that screwed up the receiver and the drive motor wouldn't stop. Took a while to figure that one out.


The transmitter, a battery and the charger.


Built by Tamas Szalay