Depth control ideas

I am thinking about an ROV to build around the T100 thrusters I am waiting for.

I don’t much like the idea of a vertical thruster, since that eats battery power constantly to stay submerged. I have been thinking about a ballast tank idea. I would build a cylinder with a piston and O-ring, or maybe even use a commercial air cylinder. A small electric motor can move the piston in and out, pulling in or expelling water.

Will I be able to control this to hold a depth? To hover at one depth Id have to make the ROV essentially neutrally buoyant at that depth, would that make it to easy to float around by small water flow?


Since its now winter here in Maine I cant do much testing for a few months, but Id like to get started building so I can launch in Spring.

Hey Ron,

I think the vertical thruster could work alright and waste relatively little energy if you can manage to make the ROV neutrally buoyant. A ballast tank would definitely give you the most precise control though and would be more resistant to changes in weight if you add anything to the ROV.

I’ve never built one so I’m not sure where to get the right hardware, but a piston and electric motor should work well.

In terms of control there would be lots of options. You could make it completely manual and control it by turning the motor on and off. You could calibrate it to three positions such as negative, neutral, and positive and switch between those. The best option would be integrating a depth sensor and using a feedback controller to adjust the buoyancy to control the depth precisely. What degree of control would you want?

Have you looked into components any more? This would be a really interesting project and probably useful to a lot of builders.


Hi, we have some experience with ROVs, having the ROV neutrally buoyant and having a vertical thruster (or three) works fine. You do need to make sure the ROV is non compressible…if you use something like foam for the flotation, it will compress as the ROV dives, making it less buoyant.

Making incompressible syntactic foam for flotation isn’t very difficult, just add some microspheres to resin.

We have some info about this on Typewriter Repairmen 2011 NURC Robot

Hi Jim,

The microsphere syntactic foam idea is really interesting. Any idea what the density of the material was or what sort of depths it could handle?


The density of the foam we made was about half of water, maybe a little more. The depths it can handle depend on the materials you use, there are different quality of glass microspheres available, and also there are phenolic ones. It will work to very deep water, if you use good materials and do it all right. They also go by the name microballoons. This stuff is sold by places that provide supplies for boats and aircraft, probably some other sources too. you can use either polyester resin or epoxy resin. For most ROVs you should be able to get by with polyester and phenolic. It will take some experimenting to see how much of each component to use, I think we got best results with about 3:1 volume of filler to resin. Mix the resin first, then add some filler, mix thoroughly, add more filler, etc till it’s done. Beware the set time of the resin. Mold making is also a fun challenge, it will probably take a few tries to figure it all out.


Hi Ron,

I mentioned The Sub Committee ( to Rusty as a resource for buoyancy control, especially their R/C Submarine vendor list (




That’s really interesting. Half density of water (0.5 specific gravity) is pretty good. I’ve looked into syntactic foam in the past - never thought of making it.


I have been looking at making basically an emergency blow system for a ROV that I am prototyping. Since I wish to go down to about 200 feet, I plan on having the system slightly negative to neutral buoyant to begin with and use a flood / blow tank to submerge the unit. The idea with the tank is it will open and flood to seawater when you want to dive and maintain close to neutral so there is little vertical thrust requirements and in the event of loss of communications or flood detection in the battery / electronics area … the system will close the vent on the tank and blow out all of the water forcing it to the surface like a cork.

Granted, this is not going to be cheap but I have sourced some air cylinders and electrical valve systems that I intend on playing with.



Sounds like a cool project. I’d be very interested to see how you implement the tank and pumping water in and out. Keep us posted.



I don’t plan on “pumping” the water in and out. On a real submarine, our ballast tanks have “vent” values on the top of the tanks. So if your tanks are currently full of air because you came to the surface and did either a low or high pressure blow on them, the vent valves when opened will allow air to escape while water floods in from open grates on the bottom of the tanks. Since I don’t wish to get as crazy as a real ballast system is, I plan on using a triple position electrically operated valve to either be in the vent / closed / blow position. The onboard air would be in a pressure cylinder and will be a possible one shot deal depending on the depth etc. that you operate it at.

Once you do a full or partial blow of the tank, you are loosing the air pressure etc. I wanted to go with a 250 psi system to make sure that I could overcome pressure down to maximum operating depth that the ROV would survive too. I have seen the plastic shells that are glued together on the OpenRov project and I don’t care for that at all. I will more than likely go with a full seal system on anything that comes apart and I might use epoxy casted in place windows for any optical sensors etc. One of the side benefits of my real job is having access to lots of toys. We build electronic enclosures and cable systems that we test to 2000 psi so I have had plenty of years messing with stuff that works.

We really want to have manipulator arms on our platform and the plan is also to make it so a diver can control it locally while on a wreck site to search inside where it could be to hazardous to go into. The great part for our project is I have a very large pressure tank at work that I can hydro test the platform to upwards of 1000 psi to make sure this puppy doesn’t die at the design depths we want to go to. The tank has a 48" ID and is 6 feet deep so we got plenty of play room. We also have a jib crane to load the unit into the tank with.

When we get to the point of building / testing the ballast system, I can post the info for it. Our game plan is to put a test weight system with the ballast and a timer control into one of the pressure tanks and let it sink to the bottom etc. and close the system up and run it to maximum depth pressure … and wait for the timer to trigger the blow system. I should hear it impact the tank lid when it shoots to the top of the tank while under pressure.




Okay, I understand now.

Have you thought of using something like paintball CO2 cartridges to supply the pressure? I know they operate at fairly high pressures and they are really compact. Just a thought.

I’m jealous that you have access to a 1000 psi pressure tank! That’s incredibly helpful for designing and testing this kind of stuff.


One thing I don’t like about CO2 is the adiabatic cooling and the tendency for frosting to occur. I would more than likely use an inert gas or dry nitrogen to fill my gas containers with. A rapid release of 250 psi might not seem like much but any moisture in the lines can cause ice formation. I will still take a look at those cartridges and see how they are designed. Who knows it might work well using those with a different gas. I just remember the old style one shot cartridges that you had when I was a kid (Think of Noah’s Ark time frame). We want ours to be reusable.<b></b>


@Harold: Do you plan to place the gas container inside the main watertight container of the rov? I just imagined a broken valve and the high pressure opening it from inside.

John, all “gas” systems would be outside of the pressure hull. If nothing else, the various tanks, pressure lines etc. all contribute to water displacement. If I stuff everything inside of the hull then I am fighting a loosing battle with buoyancy.

I don’t like the idea of all the foam floats that I see people using. We used to make what we called Barbie Playhouse coffee cup sets by putting foam cups in a net bag, inside our pressure tank, while pressure testing cables to 1000 psi and you would have all of these pretty little foam cups when we opened the tank up. Then again, I want my system to be able to go down to at least 1000 - 2000 feet. If you can’t achieve buoyancy using water displacement and shifting weight etc. you are pretty much screwed.

We are going to take some stuff out this summer and running to 200 feet with no scare at all. Just testing out proof of concept items first. Once we start playing below that depth, I will connect cameras and lights, which I want to make some too, inside our tank and watch the vehicle under pressure while we test stuff via a tether through the pressure tank. I would really love to do a tetherless system with an inertial navigation system. That is still dream stage, but it will happen.


@Harold Could you list the type of electric valve you plan on using? I want to use an adjustable buoyancy cylinder for my project and haven’t been able to find the valve that will do what I want. I too want a bi directional valve with an all closed position, but the solenoid valves I’ve seen don’t do quite what I want.

Also, they need to run on 12v so that might be an issue.

Any help is appreciated.

@Ben W, what exactly do you want for your valves. I haven’t been working on this in a while; however, I have a source for proportional control valves that will work with Seawater to pressures up to 320 Bar. That companies name is Hauhinco. If you want lower pressure proportional control valves … I use a company called Asco for controlling air up to 230 psi.

One thing about these valves is they are not small. Even the Asco air valves are not tiny so I don’t know what your space requirements are going to be. The particular series of valves I am using on my pressure tank are 24 VDC and I am using 0-10 VDC for scaled control. They make them in a lot of varieties so you would have to search on their sites.

Sorry I don’t have a specific number for you.

@Harold Thank you for the info. I myself don’t know the vehicle dimensions quite just yet since their are a number of things that plan to be incorporated. Right now I am testing proof of concept to see if it will work for my task. I am looking to 1) allow a CO2 cartridge to expel the water out of a flooded cylinder, 2) maintain a given quantity of air/water in the cylinder (closed system), and 3) allow water back in to the system. I am imagining using to electric valves. One, having the three positions which will allow flow of CO2 into the cylinder, allow water to flow into the cylinder or block both from flowing. On the other side of the cylinder i would have one other valve that operates either on or off. This would allow the water to be forced out when the CO2 cartridge is engaged. From what I’ve seen, the electric solenoid valves function in obscure ways mainly because they are used to power pistons so it doesn’t seem like they are designed to do what I am considering. I will look through the company websites you have listed and see what will turn up. Any other advice is greatly appreciated.

Ben, please check your profile. I sent you a friend request so you can view my private e-mail. It would be faster to work on a system via e-mail than forum post; however, it would be nice to post updates on a system etc. so others can get some use out of it.