AUV Design - Mechanical/Electrical Status Update (04/18/2012)
After multiple failed motor drivers, I finally convinced my teammates to allow me to use the proper wire schematic to connect the thruster-dedicated Arduino Uno to the motor drivers--a wiring scheme that I was taught and had used successfully on multiple occasions (including when testing one of the thursters several months earlier). I ordered three new L298 Dual H-Bridge motor drivers, cleanly/professionally soldered all of the components to them in the proper locations/orientations, developed three new colored, 6-wire, ribbon cables (i.e. cut-to-size, and soldered right-angle male pins to the ends so they could conveniently plug into the female sockets on the respective motor driver/Arduino Uno), and wired them properly (ensuring that the three paired timers on the PWM lines were in correspondence with the three thruster pairs). Ultimately, through all the presentations, reports, project, and other assignments, Hang and I were able to run successful air tests of both the thruster code and the mechanical subsystems code (with the newly-integrated marker dropper actuation). In addition to theorizing that one of the PWM pins (i.e. PWM5) on the original thruster-dedicated Arduino Uno was malfunctioning (verified by Hang using an oscilloscope), I also came up with the suggestion to have the mechanical subsystems code delay upon initialization such that upon conclusion of the thruster/general maneuverability code, it would actuate the solenoids and servo motor in sequence, thus actuating both of the torpedo launchers, the grasp/release mechanism, and the marker dropper. This method enabled us to avoid having to run one code, remove the AUV from the water, take off one of the end caps, switch the USB to the other Arduino Board, and then upload and run the second code--an inconvenient methodology.
Ultimately, the AUV successfully worked on Friday during our final hardware demo--the vehicle moved around at a sustained fully submerged depth, then launched the torpedoes about 6 feet (very slight modifications can be made if necessary to achieve even further distance and accuracy, including simply increasing the operation pressure to about 125 psi), actuated the grasp/release mechanism to both close (grasp), and then open (release), and the actuated the servo motor on the marker dropper mechanism to run through a sequence of rotations in order to successfully drop both of the markers upon command.
Photos/Video of the AUV at the ME Open House and at the FSU Morcom Aquatics Center final hardware demo (click on the link on the bottom left) can be seen in the photos/videos section of our website. Furthermore, the Final ME Presentation can be found in the documents section of the website as well:
http://www.eng.fsu.edu/me/senior_design/2012/team24/
A couple final mechanical features that were also made last week leading up to this demo were the manufacture and implementation of conductive plates (i.e. heat sinks) for the motor drivers, which screw directly into the back of the L298 chips and conduct directly to the top of the aluminum electronics platform (thus completing the conductive thermal network), stainless steel balancing plates, which are bolted to to the front of the bottom face of the AUV (on opposite sides of the camera enclosures) in order to neutralize the natural pitch angle to zero and increase the net density of the system slightly (to the desired design density), and the drilling of a pattern of relief holes in the cast acrylic torpedo cannons to enable rapid pressure equilibration, and thus drag reduction, during launch (this proved to be successful).
Currently, the vehicle is 100% mechanically complete, and the reproduced voltage regulator boards (my backup regulator was used during the hardware demo) should be re-integrated by the end of the week. Thus, the next phase of the AUV development will be the electrical/software integration of the IMU and pressure transducer, as well as installation of the right-angle USB cables and adapter in order to connect the Zotac to both Arduino Uno boards, as well as to the IMU and the two web cameras. This will enable codes to be uploaded directly to the Zotac--the central brain or heart of the AUV. Hang and Ryan will need to complete these tasks, and then successfully integrate/modify their mission control programs to enable the AUV to physically complete the replicated initial competition tasks. They will then need to derive programs for the "Kill Caesar" and "Gladiator Bins" tasks. I will attempt to develop PID control algorithms in order to maintain constant depths, velocities, rotate to desired relative angles, etc. over the summer--algorithms that would enable the experimental tuning of the PID gains (or at least PD gains).
If, and only if, these remaining tasks are completed I would be proponent of going to competition in San Diego, CA this July with a highly respectable product originating essentially from scratch, and from a first-year team. Only time will tell.
-Eric Sloan (ME Project Manager, Senior Design Team 4 - 15th Annual AUVSI Robosub Competition)
