My first prototype is a positive buoyant ROV that uses four T100 thrusters to create ballast and directional control. The four thrusters articulate via HiTec servos housed in a watertight NEMA enclosure. I’m using an Arduino MEGA2560 microcontroller, with on board accelerometer, Bar30 pressure sensor, and water sensors. My controller is dual joystick control for forward/reverse, left/right, cw/ccw rotation, and up/down. I also have an lcd screen and multiple pots for feedback and adjustments on the fly.
Most of my testing has been in a trash can, but I’ve finally gotten some pool time (see video link). I’m using the accelerometer to try to maintain level while changing directions. I’m hoping that once I figure out the parameters I need for responsive level and depth correction, I can add more buoyancy, increasing my thruster output and maneuverability.
My tether is a relatively short and heavy 25 wire cable. I found it’s just as easy to tie it to the bottom and let it sink versus trying to make it float. I don’t expect the enclosure to go much deeper that 15ft so I didn’t bother with a longer cable. This is more of a proof of concept but I would appreciate any and all input.
The controller has two XY joysticks (fwd/rev, left/right, cw/ccw, up/dn), two rocker switches, 7 potentiometers (ballast, speed angle, joystick zeros, accelerometer gain, etc), and an LCD screen. Since I was only going with a 20ft tether, the 25 wire seemed a simple way to communicate. Once I move to a hull that is capable of more depth and define my running perameters, I will likely move to a lighter tether.
The forward speed is relatively slow at this point because it’s directly related to the ballast thrust required. I need to get a better handle on the stability, then I can add buoyancy foam to increase the required thruster speeds, thus increasing the travel velocity.
I attached the thrusters to the shafts using hub horns from ServoCity. I think they are similar to Actobotics. I did run into an issue attaching to the shaft with a single screw at the center (without spline). This is not a very robust design because it is difficult to lock in place and still allow disassembly. Make sure you have some type of positive rotation lock between the hubs and the shaft.
ServoCity is like a big daddy to Actobotics so Im glad we talking about the same source, Ive looked in to Actobotics hubs and couplers before but got turned down due to different Mounting Hole Spacing , BR thrusters have 0.750 and Actobotics 0.770 , so slightly different in hole patterns and might not line up properly in addition the screws are not the same, did you do any modifications to thrustrers or hub ?If your hubs have a good match I would use them in conjunction with dual pinch bolt clamping hub for best grip.
Check my video below ( please ignore the issue with dual servo control, its been sorted by now)
This is a very interesting idea. Can we apply this to bigger ROVs maybe? Not having separate thrusters for the heave and yaw movement. If we can attach a servo motor that can make the thrusters rotate 45degrees in each direction so they can act as both vertical and horizontal thruster, thus decreasing the amount of overall thrusters that we have to use.
I am sorry if this question may sound stupid but I am newbie to ROV design and still have a lot to learn.
It’s possible to apply to bigger ROVs, but it does depend somewhat on how large you go. I would imagine the main issues to be in
keeping the shaft seals waterproof
ensuring the shafts and motors can handle the bending and shear forces from the thrusters when they’re turned on
controlling motion (an autopilot program like ArduSub assumes the thrusters are rigidly positioned relative to the ROV, which makes control a bit easier than if you need to take dynamic thruster orientation into account)
There’s also the question of what exactly you save/what the goal is. If the combination of the internal motors and waterproof couplings and the extra housing they’re in is cheaper than just additional thrusters then there’s a benefit, assuming the control doesn’t become too complicated, but it’s a situation-dependent tradeoff rather than clearly better one way or the other
