All… I have been scouring the net (and these forums) for a solution to fouling on a underwater camera housing we have built using a 4" body tube and a IP camera. We are looking to leave it in a freshwater lake for a few month, in about 10’ of clear water. Typically we have to go clean the lens ever week, else the slime buildup really destroys the video quality. Has anyone made a wiper system that can work on one of the clear end caps?
Why not mount a mineral-oil-filled servo alongside the tube, with whatever wiper would work best on the servo arm. An arduino could be programmed to make the servo scrub the lens for 10 seconds once a day, always returning the wiper arm to a position not blocking the lens.
I looked for mineral oil filled servos, and couldn’t find anything. Could I just fill a waterproof servo with oil and call it a day?
Yup yup. If you were going to great depths i’d suggest small aquarium air line plumbed into all your servos, fed by a pressure-compensated reservoir.
But for a shallow lake, just disassemble a common servo, coat the controller board in epoxy resin fiberglass thinned with denatured alcohol, then drill a hole in the case wherever it won’t mess with the internal workings, reassemble, then use a syringe to squirt mineral oil into the servo. Seal with 100% silicone.
After a few months/years the servo may glitch due to carbon dust being produced by the internal resistive potentiometer that provides positional feedback to the controller board. But since you used a cheap servo and some mineral oil, just swap it out.
That is quite a thought… I have a bunch of waterproof servos already lying around, and Arduino equivalent from other projects.
In theory, I could just submerge the entire servo any small container of oil, and then disassemble and reassemble.
Even though the servos are listed as waterproof, I doubt they have any type of pressure rating… But if the internals were filled with oil, then in theory there would hardly be a pressure differential.
We created a underwater camera that we could deploy at different locations in this late, and as long as it receive 12 volt power, it would record everything it’s all to a SD card.
Instead of configuring the camera for Poe, we just fed 12-volt down two of the tether cables pairs, and used the other two for data like normal for a ethernet connection.
This allowed us to power the camera through a solar battery setup, but also connect our laptop into it and pull data directly from the card without having to remove the camera.
The video footage was incredibly popular, and we are using the setup for a case study for underwater habitats, and plan on streaming live to the internet.
Exactly. People keep asking “How would i seal the shaft?”, and don’t seem to understand that with a non-conductive fluid such as mineral oil filling all the internal spaces, there is no empty space for water to force it’s way INTO. The reason i seal the electronic board in thinned epoxy is so that IF water did make it’s way inside, it wouldn’t be able to short out the servo. Especially highly-conductive/corrosive salt water.
I’ve been working on a fairly similar set-up, thinking about using a Raspberry Pi Zero with Arducam and the LED module from a dive light for night. I’ve been planning on diving in order to clean things, since we figured that a battery pack would need swapping out every week or so and would want to recover the footage, but if anyone figures out a good set-up for doing the cleaning and/or running power down (or wants to collaborate or putting together a design like that), I’d definitely be interested.
What type of water, and how deep? When I deisgned our camera system, there were a lot of things that I needed to take into account. Heat was one of them. Although it was going into the bottom of a freshwater lake, any electronics inside the housing that could produce heat, would have to be dealt with. We went with the clear acrylic tube instead of the metal, which would do a lot better job of removing heat.
For power and data, the setup was a little more complex. A 50 watt solar panel was mounted over a small bucket that housed electronics. This could be configured to be staked in shallow water, or set up on a floating buoy made out of laminated 1" building foam. A solar charge converter took care of power management, and we went with a small 12v lead acid batter like you would see in a burglar alarm system.
12V power was shoved down the tether cable using 4 of the 8 conductors. From there it powered a Hikvision outdoor camera. The last 4 conductors were used for Ethernet data, and plugged into the Ethernet port of the camera. On the dry side we could simply plug in our laptop and have a direct connection with the camera.
Some things to consider…
If this is going into salt water, you will have corrosion issues. The housing parts are made of aluminum, and the bolts that hold it together are stainless. With no zinc anode to sacrifice, the dissimilar metals will produce a small electrical charge and begin corroding the weaker metal.
Plan for all components to get wet. Even on a good day, just the humidity in the air around a marine environment can cause problems long term.
Lights take a lot of power… and I mean A LOT. We turned the IR off on our camera, and actually shut it completely off at night to save power. Even with a large solar panel, we had to account for atleast 2 days of no charge for power sizing. If there is a front pushing through, or a storm and it’s cloudy, we get no power.
Lastly, plan on a system that you can’t service for a week or two. Crappy weather kept us at the dock more than once.
Land Based Mode
Waters edge mode with solar panel removed
Solar panel removed to get to connector to put in laptop for live viewing.
Although I don’t have a good picture of the camera right now, it is sitting on the far left of the table. The legs are PVC pipe and filled with lead to weigh them down.
I’m looking at salt water, likely ranging from 30-50’ depth (I’m hoping to set up the camera in a kelp forest off the central coast CA). Seeing you use the tether for a solar connection is really neat, and definitely something I’d be interested in giving a shot. Your comments on dealing with heat dissipation, the dissimilar metal corrosion with the housing screws, and frequency of servicing are also really useful tips as well.
My current set-up is set on a time-lapse, so aside from a few seconds every ten minutes or so, the lights should be off, and the battery draw of the Pi Zero is low enough that even with a 10000 mAh @5V, I’m getting a little over 3 days of power. Unfortunately, I’m away from the hardware today, but will post pictures when I can.
Since the Pi Zero operates with a 5V in, I’m going to need to manage power–I was thinking about picking up one of the “platters” and seeing how it works. With any luck, a compatible power bank with voltage converter should be enough to run an oil-filled servo for a wiper (as suggested above) would be the way to go.
Get a 2 pack of 12vdc 10ah SLA batteries off Amazon for $40 or so, run them in parallel with a buck converter.
Shameless plug: my company, Blue Trail Engineering, sells truly waterproof servos rated to 100 m depth. They would be perfect for this.