Gas station without pumps

2011 February 13

Waterproofing cameras for underwater ROVs

Yesterday, members of the Robotics Club at my son’s high school went down to Monterey Peninsula College for a camera-waterproofing workshop.  The cameras are  for the underwater remotely-operated vehicle contest coming up in April.  I blogged about an earlier workshop in which my son played with building an ROV—the cameras are for such a vehicle. [Note: this workshop was not an official activity of the Robotics Club, but just a couple of the members going to the workshop with their dads.  Official activities have to have one of the high school teachers present at all times.]

There are now 13 “Ranger” teams registered (who have to use cameras to navigate their vehicles) and 28  “Scout” teams (who can watch their vehicles from the pool edge and so build simpler ROVs) for the Monterey Bay regional.  The Ranger teams are mostly high school teams, and the Scout teams mostly middle school teams.  Some middle schools have 3 teams entered, but the high schools seem to be running one each.  The high school teams are coming from as far away as Stockton and Santa Maria (each about 160 miles from Monterey by car), so our little commute of 50 miles seems small.  None of the distant teams came for the camera-waterproofing workshop though—only 3 or 4 of the more local Ranger teams were there.

The whole contest started in Monterey in 2002, but has since grown to 20 regional competitions feeding into the international competition.  There are over 200 teams registered in the various regional competitions.  The contest is actually international (one regional is in Hong Kong and one is in Japan and there is a team registered from Egypt), but most of the teams are still from the US.

The camera waterproofing turns out to be surprisingly easy (though I suspect that there was a fair amount of trial and error in the development of the procedure).  There are full instructions posted on-line, so I’ll only include a few photos here of some high points.

They start with a cheap video camera (the Anaconda Color model SC18A camera with a 60-foot cord). I would have preferred starting with a USB camera (since we have laptops, but not old TV monitors with analog inputs), but the workshop cost only $25 per camera (including the camera, superglue, cups, epoxy, … ), which is an unbeatable price. There is a subsidy involved since the camera alone costs $50 on the web. I think that there is some difficulty finding USB cameras with 60-foot cords also, and we really don’t want to have to waterproof USB connections.

The first step is to remove the camera from the housing, which requires destroying the housing, since the camera is glued in.

Clipping the housing off the camera

The housing that the camera comes with needs to be removed. Some parts can be unscrewed, but part of the case has to be clipped off with diagonal cutters.

extracted camera

The camera, after the case has been removed, is quite small, and has rather fragile leads. One lead broke on one of the cameras, and the student had to resolder it. This was not difficult, except for the minor problem that the classroom was part of an automotive workshop and the only available soldering irons had rather large tips.

The next step is to fix the focus of the camera, since it can’t be adjusted after waterproofing.

locking the focus with superglue

After the camera has been focused at the desired distance, the focus is locked by adding a drop of superglue to the focus ring.

The camera is glued lens-down in the bottom of clear plastic container.

Gluing in the camera

The now fixed-focus camera is glued to the bottom of a clear plastic container using superglue in a ring around the lens. The superglue forms a seal keeping the epoxy used for the waterproofing from getting in front of the lens.

A slow-curing epoxy (EnviroTex Lite pour-on high gloss finish) is used to pot the camera. Faster epoxies could be used, but the exothermic reaction of the epoxies can overheat and damage the cameras, so a slow epoxy is safer.

mixing the epoxy

Mixing the epoxy with a little food coloring, so that the cameras can be color coded.

Because the epoxy takes many hours to set, the workshop organizers had us leave the cameras in the classroom and will mail us the waterproofed cameras in a few days.

Only one boy at the workshop managed to superglue his fingers together, and they had some acetone available to release him. (Luckily for our pride, it wasn’t one of our club’s members.)  The workshop was fun for the students and they got a lot of good advice on ROV building from Jeremy Herzberg, a technician at Monterey Peninsula College who has been involved in the underwater ROV competition for several years.


  1. Neat. Makes me wonder why this isn’t in the core curriculum. Will you be using a video capture card to monitor the analog video on a laptop?

    If you are interested in using a USB camera, have you considered a USB-to-Ethernet converter? It would increase the required volume of the waterproof container, but avoid the length restriction of USB cables (beware: some devices that convert from USB to Ethernet-style connectors and cables do not change the encoding; they are merely connector-adapters, and may not solve the length issue. They may also reduce the transmission speed). Since the camera behaves as though it is connected locally via USB, you retain the ability to change the focus, as well as any other parameters the manufacturer makes available.

    Comment by Mylene — 2011 February 13 @ 13:18 | Reply

    • The students have not yet decided how to handle the analog video signal. That will probably be brought up at the next meeting of the club. Since we have now 2 cameras with analog video, we will not be looking at USB cameras this year. Perhaps next year.

      I see that “Sabrent USB Extension via Cat5e Adapter” provides a USB connection for up to 150 feet of cable, rather than the usual 5 meters.

      Comment by gasstationwithoutpumps — 2011 February 13 @ 15:13 | Reply

      • Yes — that looks to be of the type that patches USB signals to a twisted-pair cable, without changing the signaling protocol (basically a USB amplifier). Even if the distance is sufficient, watch out for reduction in transmission speed (on the order of USB 1.0 instead of 2.0). Sounds like it won’t be an issue though, if USB isn’t in the cards for this year. Good luck!

        Comment by Mylene — 2011 February 13 @ 17:01 | Reply

  2. As my mentor (whom I did not listen to nearly enough, Sorry Frank B.) repeatedly said keep it simple.
    The anaconda waterproofing works well and has hundreds of cameras behind it. The Move to USB must be very carefully tested. In the event you blow up, flood, muck up the camera you will lose more than $100 in equipment. This is not to say there are not potentially some serious advantages. For example this may be more resistant to motor noise?
    If the goal is to use a better camera-this may be one way to do it.
    If the goal is to bring the video feed into the computer, could that be done safely with a composite video to USB adapter on the surface?
    If the goal is to reduce noise, try better wire?

    So first step-define your goal
    Second step experiment

    That being said, experiment, experiment, experiment. You rarely ever know exactly what you will find, but the contest (I belive) is about finding those people who are willing to take the trip.

    The only wrong way is if somone says this is the only way :)

    Jeremy R. Hertzberg, BSCMPE

    Comment by Jeremy R. Hertzberg, BS CMPE — 2011 March 2 @ 18:45 | Reply

    • I concur that USB cameras would be a much more expensive way to go.

      We have already ordered some video-to-USB adapters (EasyCAP DC60 – USB 2.0) that cost under $10 each, and we’ll see how they work—at least there was freeware available for running them from a Mac. I don’t know if we’ll need 2 laptops, or if one can run both cameras.

      Since it is all coming out of the pockets of two of the parents (me and one other), we are trying to keep the parts cost down. So far the wire and PVC parts have actually come to the most (things like expensive stainless steel U-bolts for mounting the motors), $100. We’re up to about $70 for vision (cameras and adapters, not counting laptops), $75 for 4 bilge pump motors, $20 for propeller adapters (which we may not use, if the new propellers work better than the ones tested so far), $20 for propellers to test. I figure we have another $25 for switches, fuses, and the control board, and another $20 for more PVC. Not to mention the props, which they forgot to get the parts for. So the whole thing is running between $300 and $400.

      I’m thinking of loaning them a 6.6Ah LiFePO4 battery I have for a project that is not going to be completed in the near future. They should be able to run 2 motors at a time off the battery, and it is much lighter and more rechargeable than lead-acid battery.

      I plan to put up some pictures, once the students get some building done. Not sure when that will happen though, as these kids have busy schedules.

      Comment by gasstationwithoutpumps — 2011 March 2 @ 20:27 | Reply

  3. Sigh, the (EasyCAP DC60 – USB 2.0) adapters arrived yesterday, but the combination of Anaconda camera, EasyCAP DC60, EasyCap Viewer, and MacBook Pro did not work. I don’t know where the problem is, and I don’t have the tools to debug. I’d need one or more of the following: another composite video source, a monitor that took composite input from a phono plug, and a Windows computer that I could install the EasyCap drivers in. Lacking all of those, I’m not sure what my best way forward is.

    I contacted the developer of the EasyCap Viewer software, and he was mystified. He did not that the Version 10 of the EasyCAP DC60 is older than any other he has seen—every one else seems to have Version 11. Perhaps the $9 adapters were not such a deal after all.

    Comment by gasstationwithoutpumps — 2011 March 4 @ 17:51 | Reply

    • Bummer. Wish I had some useful suggestions, but the products I’ve got in mind are unreasonably priced for this application (pinnacle’s video capture product, for example, is in the $100 range). At the same time, I take Jeremy’s point, above — if you’re going to buy something expensive, better that it’s at the surface, not in the water. In your shoes I’d be tempted to borrow a Windows computer, a VCR (source of analog output) or an old TV with a composite input for troubleshooting. You mention that there are only two parents contributing funds… might there be a parent who can’t contribute money but could loan one of those items, either for troubleshooting or for the length of the project? If you find a reasonably-priced video capture product, I’d be curious to know.

      Comment by Mylene — 2011 March 5 @ 09:31 | Reply

      • Well, there are only 3 students currently involved, so there isn’t a lot of extra parent depth to go to for funding.

        I’m pretty sure that the cameras work (they were tested at the waterproofing workshop), and that the problem is in the video capture device or the software. The most likely is the software, since the developer admits he hasn’t seen version 10 of the hardware before, and version changes in the hardware are likely to result in different demands on the driver. That means we might be able to work with a Windows computer. I will find out if any family in the club uses Windows.

        Comment by gasstationwithoutpumps — 2011 March 5 @ 11:05 | Reply

        • just started first rov some great ideas. camera the big one

          Comment by chris — 2013 July 22 @ 17:44 | Reply

  4. […] Waterproofing cameras for underwater ROVs ( […]

    Pingback by Underwater ROV contest « Gas station without pumps — 2011 April 3 @ 08:58 | Reply

  5. […] Waterproofing cameras for underwater ROVs ( […]

    Pingback by Underwater ROV again this year « Gas station without pumps — 2011 December 22 @ 21:37 | Reply

  6. […] Waterproofing cameras for underwater ROVs […]

    Pingback by Blog year in review « Gas station without pumps — 2012 January 1 @ 14:17 | Reply

  7. […] Waterproofing cameras for underwater ROVs […]

    Pingback by Second Blogoversary « Gas station without pumps — 2012 June 2 @ 18:15 | Reply

  8. […] Waterproofing cameras for underwater ROVs […]

    Pingback by 2012 in review « Gas station without pumps — 2012 December 31 @ 11:17 | Reply

  9. […] Waterproofing cameras for underwater ROVs […]

    Pingback by Post 1024 | Gas station without pumps — 2013 March 23 @ 19:59 | Reply

  10. […] Waterproofing cameras for underwater ROVs […]

    Pingback by Blogoversary 3 | Gas station without pumps — 2013 June 1 @ 20:00 | Reply

  11. […] We would use Arduinos or KL25Z boards, rather than Lego NXT bricks, and waterproofing their designs could be a part of the engineering. Using a drybox and a connector should not be too difficult. The robotics club has used IP68 Sealcon strain reliefs from for cables and Buccaneer mini IP68 connectors for disconnectable connections, both successfully. Some parts could be potted in epoxy, like the cameras for the underwater robot. […]

    Pingback by Electronic sensors for water quality | Gas station without pumps — 2013 September 22 @ 16:19 | Reply

RSS feed for comments on this post. TrackBack URI

Leave a Reply

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

You are commenting using your 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 )

Connecting to %s

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

%d bloggers like this: