9. It’s impossible to read what the real time values are
for different variables once the Quad_ROV is in the water.
This can make development and troubleshooting difficult.
While these values could be sent up the tether as serial
data, this would slow the processor to an unacceptable
level. The answer might be to add an SD card plug to the
controller circuitry. Various data could then be written to
the SD card in real time and analyzed after a test run. This
would require use of the SPI pins, and pin assignments on
the Teensy 3.1 would need to be changed.
10. Twenty centimeter long header pin jumpers were
used; using shorter ones would make for less of a wire
tangle when they are installed.
11. There is not a light for underwater illumination on
this prototype, but it could be easily added. A waterproof
light — like that shown in my March 2016 SERVO article —
would have its power leads fed into the box and directly
connected to the 12 volt power line.
12. While developing the controller software, it was
necessary to reprogram the Teensy 3.1 many times. The six
inch PCB used for development has the Teensy 3.1 in a
position where it is impossible to plug and unplug the micro
USB cord used for programming. Instead, a six inch USB
extension cable was permanently plugged into the Teensy
3.1 and brought out to the top of the box to make
connections to the programming cable easy.
13. While it’s convenient when testing to be able to
strap the ballast onto the Cantex box with bungee cords,
this isn’t an ideal position. This positions the center of
gravity quite low, and when a pitch or roll is required, it
strains the motors attempting to attain the desired angle. A
better placement might be between the Cantex box and
the PVC mounting sheet that is attached to the arms. Once
in place, this would make entering the Cantex box easier, as
well as the ballast would not need to be removed. SV
SERVO 01.2018 41