The intervehicle communication capstone project was created to address the Underwater Robotics Team’s need for a functional communication system within a tight budget. The project focused on the development of an acoustic modem, capable of receiving and transmitting pressure wave signals. The team consisted of me (a mechanical engineer), three electrical engineers and a software engineer. The electrical engineers developed a custom PCB to drive the modem and filter received signals while I developed a custom transducer.
The need for the transducer came from a need for transmission range at a below-market price. The guiding principle of the design was to ensure that it was very robust – such that it could operate for many years to come. Additionally, the device needed to be easy to maintain, have a range that could cover an Olympic-sized swimming pool and have a low production cost.
Transducer Theory
The transducer operates on the same principle as a typical loudspeaker. A coil, within a strong magnetic field, is subjected to an oscillating electric current causing it to vibrate at the driving frequency. The coil then drives the speaker head, creating a pressure wave. The farther the coil moves, the more intense the pressure wave.
Electrically, the coil behaves as an inductor and resistor in series. The resistance and inductance – and the overall impedance – increase with the number of coils, however, so does the oscillation amplitude. Additionally, the circuit’s complex impedance depends on the oscillation frequency. The combination of these factors created a challenging optimization problem.
Waterproof Design
Inspired by the T200 thrusters – used by the Underwater Robotics Team – the transducer was designed to be waterproof without sealing. While this may seem counterintuitive, this simplified the design and eliminated the seals as a potential failure point.
The transducer was able to operate while flooded with water as it was constructed from materials – notably Samarium Cobalt magnets – that are extermely corrosion-resistant.
The flooded core design offers some additional benefits. Thermal issues, which affect the device in air, are nonexistent in the water. Additionally, since there is no pressure differential between the core and surrounding water, the device can operate at extreme depths.
Manufacturing
As the transducer was of a custom design, it needed to be manufactured entirely in-house. Most of the core components were machined using a vertical mill and lathe. The head designs were 3D printed and various print settings were tested to achieve optimal performance.
More interesting was the fabrication of the voice coil that drove the head. This coil consisted of very fine-gauge (thin) enameled wire wrapped around a thin aluminum bobbin. As the coil needed to be wrapped both tightly and uniformly, a custom coiling device was constructed. This device maintained tension on the wire and tracked the number of turns, allowing an operator to organize the wire as it was wound.
Testing and Validation
Validatation of the transducer’s design involved performing a comprehensive series of tests.
To ensure the device could withstand a harsh marine environment, the transducer was submerged in pool water for 72 hours. Although some galvanic corrosion was found, it could be easily flushed off and did not affect operation.
To ensure the device could survive operation, the device was fixed to the Talos AUV for a series of pool tests. While the device held up during operation, the coil was damaged when the lead was snagged. To address this, an updated mount was designed that featured a strain relief for the coil lead.
Finally, to verify the device’s operational depth, the transducer was sunk to the bottom of the university dive well. While the well was only 18ft deep, no drop in performance was noted.
Electronics Integration
To interface with the rest of the project, namely the electronics, a prototype was built from an off-the-shelf speaker to allow the electronics team to validate their designs while the deployment model was in production.
Additionally, empirical testing was performed to find the optimal number of coils in the transducer. More testing was then done to determine the ideal frequency of operation for the entire system.
