Hi,
I am working on my first ROV. I would like to have your advice about the design that I intend to build, because I am not sure that this is a good or feasible idea.
The ROV main body will be a traditional tube, and it will have 4 thrusters:
- 2 thrusters oriented vertically, in order to provide vertical motion and roll control (with differential thrust). They will be placed around the center of gravity of the ROV, on both side of the main body.
- 2 thrusters oriented horizontally, to provide back and forth propulsion and yaw control (with differential thrust). They will be placed at the stern of the ROV, on each side of the main body.
In order to have some pitch control over my ROV, I would like to be able to orient the back thrusters around the pitch axis. To do so, I think of two solutions:
- Link the motors to a servo inside the main body of the ROV, using a shaft that would cross the body of the ROV, perpendicularly to the main back/forth direction of the ROV.
- Put one or two servos outside the main body of the ROV, linked to the motors to orient them.
Solution 1 tempts me, but I think it is a challenge to make the shaft crossing the hull of the ROV and preserve the watertight of the main body, while allowing it to rotate around its axe. I could not find any solution for this on the forum I visited. Does anyone has an idea on how to achieve that? What depth could be reached with such a system? Do you think that it is too difficult to have a watertight connection between the motor and the inside servo?
Solution 2 is much simpler, but the downside would be the price. I found some watertight servos on the internet that rated around $100! It is too much for my budget, so I was wondering if there would be some simple way to make a standard servo watertight? Compare to solution 1, this solution seems to me safer because if the watertight servo fails, I only lose a servo, and not the full electronic onboard the main body of the ROV!
So I would appreciate to have your opinion about this design, and the best solution to make it real!
Regards,
BM
BM,
You are correct, option 1 can be tricky and carries a higher risk of repair bills. VideoRay successfully use a thruster through the housing for their vertical thruster on some models and it works good, but the seals are expensive and don’t last long (though you would get more life and cheaper seals if you didn’t have the higher velocity of the thruster). While some ROVs use tilting thrusters, this generally limits the practicality as you often want to hold a fixed pitch while flying. One can also use a mass (could be the battery) on a server or acme screw to change the center of mass, resulting in a tilt. You might consider this an option. Also, If you can put a tilt servo on the camera, that would work in most situations. On my 6 thruster, I rarely need tilt my camera more than the 45deg down/up. But I really like the stability of the 8 thruster (which you won’t have with 4 thrusters).
If cost is your focus, I would recommend 3 or 4 thrusters and just ditch the pitch control altogether. Possibly do one vertical forward and one aft and keep your center of mass well below your center of buoyancy so you don’t need roll stability.
If stability and maneuverability is the goal, 7 or 8 thrusters are pretty much required. There’s a reason almost the entire ROV world uses 6 thrusters at a minimum and 7 (3 verticals, usually one aft and 2 forward) or even 8 thrusters (1 vertical in each corner) are the preference. The stability is unmatched. It used to be that DIY ROVs focused on minimizing the number of thrusters, but that’s when we were paying $2,000 each. With the advent of a great thruster and controller for under $200 total you’ll almost certainly save money in the long run replacing servos and seals with more thrusters.
Well, even with $200 for each thruster, the prices is still an issue for me when it comes to 8 thrusters ($1600)!!! 
But thanks a lot for the comments and your advices, it really helps me a lot. I like a lot the idea of a moving battery to change the center of mass, I think it is very clever! It remembers me of this project where pitch and roll where controlled like that: OSUG: Open-Source Underwater Glider | Hackaday.io
However this solution comes with the risk to get stuck in the vertical position without possibility to come back to the horizontal! 