The ROV-project started in 2007 when I was in high school studying media and communication. I came across a great website (Homebuiltrovs.com) that describes the build of several types of ROVs, only using "garage technology". Since I didn't have access to proper machine tools at the time, this project was perfect!
The frame was made from a bottle crate that was modified by cutting off the top part. This left me with a plastic mesh with one column in each corner, perfect for mounting cameras, thrusters, etc. Buoyancy was provided by four solid plastic spheres, orginially intented as floats for fishing equipment. These floats already had mounting tabs for rope, so they were easily mounted to the frame using threaded rods.
Propulsion is provided by three modified 500GPH bilge pumps. The impeller housing was cut off and the impeller removed from its shaft. A simple coupling was then mounted to the shaft, and the propeller mounted on the coupling. The couplings were made with only a cheap drill press, so it was hard to align all the holes. The propellers were obtained from an RC hobby store. This thruster configuration gave good maneuverability, although a fourth thruster to provide lateral thrust would have made it better.
The light at the front was made from some pieces of plastic tubing laying around. It contains a 35W halogen bulb, not really enough for any practical use. Check out my 35W HID dive torch build for a more powerful torch! Located next to the light is the camera. This is a circuitboard-mounted black-and-white CCD camera, potted in a substance called Raytech Magic Gel. The tiltmotor is a geared DC-motor housed in a film canister filled with wax and parafine gel. This method for waterproofing motors is explained in the great book 'Build your own underwater robot and other wet projects' by Harry Bohm and Vickie Jensen.
Controlling of the ROV is done by a small joystick that triggers the horisontal thrusters, and a toggle switch that triggers the centre motor to ascend or descend. There are also toggle switches to control the camera tilt and manipulator.
The finished ROV.
Control panel during pool test.
Left horisontal thruster. Lead weigths were used to obtain proper trim and neutral bouyancy.
The motors are each connected to two relays where the poles have been reversed on one of the relays. Depending on which of the relays are triggered, the motor will run either forward or backward. The relays are placed in the grey project box in the aft of the ROV, which is filled with candle wax to prevent water intrusion. A problem with this metod for waterproofing is that the wax contracts upon solidifying, putting stress on the electric components. In fact, it was only after the third try everything worked after potting.
I also made a manipulator. This was made from acrylic plates and aluminium sheet metal. A geared motor turns a shaft which is connected to a screw. As the screw turns, a hinge is actuated, opening the jaws of the manipulator. The motor and gears were scavenged from an old CD-player.
ROV during pool test. The manipulator was removed during the test.
The inner workings of the manipulator.
Light, camera and manipulator. The syringe is filled with air and closed in the other end, so the piston will compress the air as the external pressure increases. By observing how much the piston has moved, one can get a crude indication of what depth one is at.