Two Generations Of Ocean-going ROV

[Eirik] wants to help inspire others to take on big projects to he sent in a link to his ROV project. He started it about one year ago and the image above shows the first generation. After the break you can see the video that the ROV captured during a couple of it’s initial voyages. They’re pretty clear and right off the bat you’ll see the little guy following a jellyfish. Like a lot of homebrew ROV’s [Eirik] is still searching for the right way to pass wires through the housing without leaks. He does okay so far, and has designed a nice cable spool for the topside tether, but some water does get in. He’s almost finished the second generation which re-designs the camera mount to aim downward so that what’s in frame is more interesting.


19 thoughts on “Two Generations Of Ocean-going ROV

  1. Waterproofing wires isn’t terribly hard to make for reasonable pressures (under 500 PSI although that’s something like 1,122 feet deep!).

    Start with a NPT nipple into the body of the robot. Then buy a NPT cap. Drill into the NPT cap to fit the wires you need to put through, then push the wire through it. Epoxy the wires in place. You will not be able to remove the wires again but if done right, you have 1″ or so thick of epoxy – which isn’t going anywhere and as long as the epoxy sticks well to the jacket material, you should be good to go. We did this with teflon coated wires and had trouble getting a good seal. PVC sticks pretty well.

    You can of course buy commercial versions of these “waterproof” wires in small to large sizes as well but the price is quite high. You can find ones suitable for vacuum, pressure, etc.

    “Cord grips” might also be helpful but only to a limited submersion range. Highly doubtful you can take those past 20 or 30 psi but that is still about 66 feet or so deep.

  2. The types of seals you would be looking for in industry are fairly pricey and rather hard to locate.

    PAVE is one of the ones I have experience with:

    These seem like they are worth checking out too:

    Another idea would be to use Swagelok type fittings (Parker’s A-Lok series, Yor-Lok, etc) as in smaller sizes they are good to 4000+ PSI and are available in stainless steel (and titanium although the price is much higher) if you need saltwater resistance. A 1/4″ OD tube to 1/4″ NPT fitting in 316 stainless steel might run you $10 or so and be good for 4000+ PSI. That would be rated to at least 8979.59 feet!

    The hardest part about that though is that you have to make a rigid tube connection somehow. For something 100+ feet long, you need to be able to spool and unspool it but it’s something to think about.

    More about epoxy sealing:

  3. while the slip ring tether is interesting, the rest of the project is a bit more questionable…

    “I was never able to make the ROV entirely watertight, and as the main chamber filled with water the ROV would begin to tip backwards.” – yay for testing! (or not)

    “A note to anyone who none the less wants to use an unsealed hull solution, any water vapour trapped in your ROV will quickly fog up the interior.” – throw some desiccant in there?

    “All told, this project set me back roughly 500 USD” – really?

    “Once everything was stuffed (literally, it was a damn tight fit) I tested to make sure none of the connections had failed. With all checks good, I could proceed to potting the entire assembly in wax.” – or you could build use gasket and build a watertight enclosure out of something other that pvc…

  4. Combine this hack with some evanescent wave coupling. Have a cable with an inductor on the end and line it up with an inductor inside the ROV. It might be easier to waterproof that way.

  5. It would be messy and not exactly elegant, but how about filling the chamber with mineral oil? It doesn’t conduct electricity, it would prevent water from entering, and you could go as deep as you wanted because it wouldn’t compress.

  6. another thing for waterproofing is using pvc tubing with the cable running through it. you would have to buy a spool or two depending on how far you want it to be able to travel. also, a couple of home made underwater microphones would be cool. just use the same method for connecting those. a short length of pvc tubing to connect them from either side of the ROV to the microphones.

  7. Hermetically sealed relays have a glass seal. I would imagine that they would implode at depth. What depth, I can’t say but probably only a few hundred feet? Plus – that gets you inside of the relay while the outside remains “wet” (and conductive). If some of those pins act as a path to ground… that may not work well. But this is about keeping water out of the submersible ile still allowing cables to enter. I don’t see how the relays accomplish that?

    Even water compresses. People say it doesn’t compress all the time but that’s not technically accurate. Underwater explosions – even techtonic events are no joke – the “throw” energy of those is…. unexpectedly large compared to what you would see in air (a compressible fluid).

    Of course it takes quite a bit – probably 10,000 PSI or more to compress more than a percent or two. But at 50,000 PSI, you are going to be somewhere north of 10% compressed.

    But that’s something near 112,245 feet deep if you convert it into depth numbers!

    The container holding the mineral oil would still be subjected to pressure from the water equally in every direction. Only now, the (presumably floppy) container would be imparting that pressure and transferring it into the internals of the submersible! You can build subs out of feet thick steel and at significant enough depths, that sub is still going to be squeezed quite a lot!

  8. Now we’re talking my profession! (I work for the Navy, Unmanned Systems).

    Anyway, anyone interested in building their own basic ROV, check this site out:

    It’s a Federal Education Outreach Program where middle to high school kids build a basic ROV from cheap parts widely available. A full kit cost $125 if you buy it from them. But they also provide a parts list with sources and model numbers as well as online build instructions.

    I’m actually working to bring the program to a local museum as a summer camp for kids this summer.

  9. If you fill the air voids with a non-conductive liquid such as mineral oil you can get by with simple o-ring seals to keep the water out. I’ve used this technique successfully with cheap LED flashlights down to around 100ft while scuba diving.


    You can buy them from mcmaster carr (among other places), they are ip68 rated, many can safely go 10s of meters underwater. These are (as far as I can tell) the best way to pass a cable from the inside of a housing. The only (minor) caveat is that the cables must be round, generally not a big deal.

    After extensive research for an underwater housing I build for a dslr, I found cable glands were really the only acceptable choice for passing cables from the main enclosure to the smaller enclosure that housed the controls (just 2 buttons that share a common), so a 3-conductor round power cable worked fine.

    If you are running cable outs for power, a similar cable to what I used would do fine, if you are running data, I would imagine a good quality cat-5 type cable would do swimmingly (HAHA clever wordplay!)

  11. Another important note on the PVC,
    some pvc is cellcore (depends on where you buy it, but the pipe will almost always be marked cellcore or foamcore if you look closely). I’m not entirely certain as to the specifics of the difference, but I know that:

    For the same diameter and wall thickness, cellcore is lighter, so it seems as though it is less dense, either that it is made from a different compound (I doubt), or it has been somehow aerated to make it less dense (there are no noticeable bubbles in the cross section).

    I’ve used solid sch40 6″ pvc to depths of 60ft with no trouble, and judging by it’s pressure ratings, I suspect my seals would fail long before the actual pipe did.

    Smaller diameter pvc pipes of the same schedule should have the same pressure handling capabilities, although perhaps slightly different since pipes are rated for internal pressure (a pipe can handle much more external pressure than internal)

  12. Why not use fiber optics to pass data through the “hull”? That way the only hole you have to make is one for the power cable. The “data window” if you will, would simple have photo-isolated ports sort of like a waterproof pegboard that the optical to electrical converters plug into.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.