High voltage ROV adventures

[Eirik] wrote in to share the build log for the third iteration of his underwater ROV. The first two project were completed and tested (you may remember reading about it back in January), but both had issues that caused general failure. Most notably, the introduction of water where he didn’t want it. But this time around he seems to have gotten everything right, successfully taking this little guy down to twenty meters without a leak.

One of the problems he had on version two was supplying electricity from the surface. He needs 12V at up to 10A, and had to use a tether made of 14 AWG to make it happen. That’s a lot of heavy wire to be hauling around and it made the ROV virtually unable to move itself. He wanted to go back to using Cat5e cable but it won’t handle that kind of current. He ended up using an inverter at the surface to up the voltage to 130V, and a switch mode supply on the ROV to get back to 12V. This caused noise on the data lines which he fixed by adding a full-wave rectifer to the inverter’s output.

The dive video after the break shows off the crystal-clear camera shots this thing can capture.



  1. brad says:

    Embedding = Broken

  2. linksaregood says:

    Thanks for putting the link to the video in the posting. It makes it so much easier for those of us who download and view it offline.

    • RobinJood says:

      I’m not being a smart ass but you know that if you’re running youtubes HTML5 videos then it’s just a matter of right clicking on the video and selecting ‘copy video url’ at any time…

  3. Dzl says:

    Nice project.
    Thanks for sharing another example that high voltage ROV opeation is possible if you know what you are doing.
    I routinely use 240V mains voltage directly (with safety relay and clamp diodes) over cat5 for small ROV projects.
    Will try the inverter approach for 12V surface/boat operation.

  4. kay says:

    This can be used with one of those ethernet-over-AC things for control and video, as well. The original OpenROV stack used that (this was before Geoff ran off with the prototype). http://www.youtube.com/watch?v=QLF_SO9Psjw Here it is.

  5. GameboyRMH says:

    The guys in the OpenROV project could use this PoE idea, they were having trouble getting enough power through the cord at the standard voltage.

    • kay says:

      See the video: I was asked by the OpenROV “founders” to design them a ROV, I did, and they started showing my prototype off telling people they made it. Then a third party decided to keep the prototype and I got fed up by the whole thing.

  6. jackkrause says:

    So since ohmic losses go as current^2, couldn’t you use 3 phase power over 3 power wires that together would require less copper than a 1 phase wire while still transmitting the same amount of power?

  7. Brian Lee says:

    I think this would be useful for fishing!

  8. Mikey says:

    Just switching to A/C in general would have fixed his “I need 10 amps and giant wires” problem, not sure why he decided to transform the voltage more than 10x as well… maybe I’m missing something.

    • Rick says:

      Yup, big time.

    • borgar says:

      I fail to see how switching to ac will reduce cabling needs what so ever. if anything the skineffect will increase your need for cabling.

      upping the voltage however(and reducing the current), will reduce cable needs significantly.

      • Ken says:

        The skin effect?


        The skin effect doesn’t even start to have any practical effect on power transfer until you start getting into the 50-100KHz range. At this point, you better be using powered core / ferrite transformers because your hysteresis losses will otherwise dominate.

    • cplamb says:

      Wouldn’t there be power loss to the water through induction?

      • Aaron W. says:

        Probably we are missing some Information that led to the high voltage decision. Otherwise even the losses of a thin wire wouldn’t matter much, as long as the electrical insulation does not block the cooling of the cable. Just compensate the losses with a higher input voltage. Power efficiency of the base system shouldn’t be of much concern.

        I am also wondering what the power consumption of the vehicle is. In a submarine you usually have the problem of too much lift (because you have to provide dry, air filled space, needed by your equiment) not to much weight, so carrying batteries for one or two hours of operation shouldn’t be a problem.

  9. Galane says:

    How to offset some of the weight of a tether.
    Make buoys with a sealed, air filled bladder inside.
    Put a hole in the outer shell so water can enter.
    As the tether descends, water pressure will collapse the air bladder, reducing its buoyancy.

    That’ll work until the air bladder is fully collapsed, so you’ll want higher air pressure for the buoys that will be pulled farther down. The shell of the buoy is to keep the air bladders from expanding too much and bursting, especially the ones that will be pulled deeper.

    Something else to look into, syntactic foam. Is that stuff available as a DIY mix and pour system?

    Urethane foams like the mix and pour types available from Smooth-On can withstand a lot of pressure and should be suitable for depths the average Joe HaD’er ROV builder will get down to.

    The syntactic foams are uncrushable, even down to the deepest part of Earth’s oceans.

  10. KE7EHA says:

    Two suggestions:

    You should look into a modern computer PSU. They use a power-factor correction circuit that generates about 400-600 VDC which is then sent through an isolated switching power converter to step down to the required voltages. May be a source of higher voltage DC when and if you need to increase the amount of power sent down the lines. Just be sure to get appropriately rated cabling.

    Secondly, are you using single ended signalling back to the surface? if so, any common mode noise will eat your lunch. If you switch to differential mode signalling for analog signals and run them over a twisted pair, the common mode noise will be readily filtered out. You can also use differential mode RS-422/485 for serial signalling.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


Get every new post delivered to your Inbox.

Join 96,725 other followers