I'm Going Under (water)

[Snarls]

Not sure where to post this. Maybe better off on the Flotetest forum, but I guess scratch building is technically correct.

I was thinking about ocean kayaking the other day (as you do), and had an idea. I am a recreational fisherman, and have always wondered what I could see under the surface of the water. So naturally I thought about putting an FPV camera on a string and sinking it down into the depths. Then I started thinking about what if I could move the camera around. What if it did not have to be dangling under the boat, but could roam around the area. Suddenly I found myself in the world of ROVs, remotely operated (underwater) vehicles.

You may have seen ROVs on various educational TV shows, where they are deployed to search shipwrecks or scan the sea floor. This is usually with expensive, industrial use ROVs, on multi-million dollar vessels. From my brief research, there is not much out there for consumer ROVs and home build ROVs. OpenROV has a DIY kit, and a 'ready to dive' product called Trident, but these are far above my budget for what I want to do. I just want something to bring along with me in the kayak, that I can then deploy and control around the depths while watching an fpv feed on the surface. So here I am, building my own ROV.

My budget is pretty much $0 since I want to reuse parts, and don't want to get fancy with the build.

Here is what I have available:
- 3 (maybe 4) cobra 2204 2300kv motors
- 2 Sunrise Cicada 20a ESCs
-1 Turnigy Plus 18A ESC
-Eachine FPV Cam
-Xioami Yi with waterproof case
-Arduino Uno and Nanos
-Wii Nunchuk, Playstation controller
-3D Printer

I want to have forward, reverse, turning, and diving/surfacing capabilities. Our RC frequencies do not go through water beyond a few feet so I am not going to use a typical RC transmitter-receiver setup. Instead control will have to be done through a tether somehow.

Let me know your thoughts, and if you've done a project like tell us how it went and if you have any advice.

 

[Hai-Lee]

You could achieve a similar result if you followed the example of the Royal Navy which uses/used helicopters to dip their Sonar transponder into the water to check an area and then fly off to the next area for a similar check and so on.

You could use a drone with floats and have your camera on a winched tether for depth control. a bi-directional DC motor and prop on a horizontal boom could be used for camera pivoting or holding station.

Your standard Radio could handle the extra requirements if it has enough channels and an video transmitter on the drone could transmit the video signal to your tablet in the canoe.

As it does not need to spend endless hours flying or pushing its way through the rather viscus water the power requirements would be moderate. If power was lost the floats would mean that the drone was recoverable as long as you do not lose sight of it totally.

Just my thoughts on the subject!

Have fun!

 

[Snarls]

Nice idea Hai-Lee! Certainly a viable option, but I want to keep pursuing a full ROV design. I'm going to use this project as an opportunity to learn about controlling brushless motors without RC equipment, and to make and program my own controller (vs mixing something into my Taranis).

 

[makattack]

Heh... you young'ns... back in the early 90's, we used to call these UUV's (unmanned underwater vehicles) or UUAV's (for the autonomous variety)

I actually prefer those terms to this whole ROV (which seems too generic and could mean air, sea, land) terminology, but then I am an old guy resistant to change.

Enjoy your project! BTW, if manual/commanded control isn't required, the easiest way to avoid a tether is going the UUAV route.

Note sure if these kits are worth it nowadays, but I recall seeing ROV ahem... UUV kits for $200 from MATE:

https://seamate.org/

 

[Snarls]

Right, I don't know why these machines have become known as ROVs, but thats the name I've come to know them as.

Those kits would be a good start if I did not have so much extra parts lying around. Funny that you bring up MATE. A from of mine is currently working on an ROV with a team from his university. They are going to bring it to a competition sponsored by MATE in Seattle in a few weeks. I've been picking his brain about his ROV build, but it is much more complex than what I am currently aiming for.

 

[PsyBorg]

You haz 3d printer...

Make a ballast tank that operates on screw drive. Think syringe to draw in or push out water to raise and lower depth then its simple matter of how complicated a drive system you want. Single motor drive is doable with rudder and elevators to steer with. Very basic stuff. Can be done with Arduino or maybe even easier with an FC and a usb link to the receiver tab so it can be controlled with a normal radio

Look up RC submarines for ideas how that screw drive ballast works. There are expensive pre made systems as well as home made ideas. This was before 3d printing so you have a huge advantage to be able to make your own..

 

[Snarls]

Here is what I have decided to go with so far. I considered a ballast system and a control surface/thrust vectoring based steering methods. Ultimately, I decided that going with three brushless motor thrusters was the best/easiest option. The ballast system is accurate to most real submarines, but I did not want to deal with compressing air and water tight seals on the intake. Similar with rudders and servo actuated thrust vectoring, I did not want to deal with keeping a water tight seal with moving components.

With three thrusters, I can put one vertically, and two in the back. The vertical one will spin accordingly to pull the craft up or down the water column. The two in the back can work in tandem to go forward or backwards, or they can use differential thrust to turn the craft left or right. Using this method of control, the only moving parts are already outside the body. The areas where wiring enters the body can be completely sealed to prevent water from coming in. I will still have to develop some watertight gaskets for the hatches that will be removable for servicing.

There will be a hatch up front to get at the FPV camera. The Xiaomi Yi will have a hatch of sorts, but it will not necessarily be watertight. The Yi will be in it's own waterproof case and removable from the body if needed. Making the compartment for the Yi watertight, while still having the lens pointing out the front and having the buttons accessible was not worth it if the Yi in it's case is already waterproof. A larger rear hatch is there to access the electrical components, which at this point is just ESCs.

Between the rear hatch and central thruster is a mounting point for the tether. I want it to be rock solid so I can pull the ROV up and out of the water by just pulling the tether. There is also a horizontal rectangular section on the front where I will install some LEDs. I'll have a way to turn on the LEDs if the lighting conditions get bad, and hopefully they will illuminate the scene, and not wash out the camera.

All bolts for the hatches/tether have to thread into nuts, but those nuts and bolt holes can not be exposed to the inside of the body. If they were, I would be concerned that water may ingress over time. So I'll have to figure out a way to manufacture the body with no bolt holes exposed to the inside, yet with nuts secured enough not to pull out under force.

A shown this design is 160 x 240 x 80 mm. Props will be DAL 5040 props cut down to 60mm.

 

[LitterBug]

Anxiously awaiting your progress! This is something I've thought about doing a few times....

Cheers!
LitterBug

 

[FoamyDM]

the 75 mhz crystals transmitters will work through the water, for un-tethered controls.
https://www.rc-submarine.com I think sell equipment just for subs. This guy is amazingly approachable and can help you do it on the cheap. the trick will be to get the ballast to put it just at the surface, so the failsafe is to surface.

i have a couple posts on my LPL flying submarine project about setting the sub part of the project up.

keep it up, are you going to 3d print the shell?

 

[Snarls]

Looks like a good resource for specialty parts. Thanks for the share. I going to keep the control tethered. I think I have an old 72mhz radio I could use, but I'm planning to dive up to 50 ft. At the very least, a tether would provide a safety link to pull the craft back up if it gets stuck/I lose control. I'm planning on 3D printing the shell in multiple pieces. Then assembling and sanding and epoxying until watertight. I want to make it neutrally buoyant so I can place it at one depth and keep it there without keeping the vertical thruster on and requiring a depth gauge.

 

[mikej997]

Just a thought, but you could use a movable ballast weight to control pitch. The mechanism could reside completely within the sealed body but allow you a certain range of nose up/down, and since you'll likely have to add weight anyway to get neutral buoyancy...

 

[Snarls]

Not a bad idea Mike. I may have to implement something like that if I find the pitching capabilities insufficient with a combination of vertical and horizontal thrust.

 

[Flying Monkey fab]

Something I noticed missing here is unless you're going to stay very close to the surface you're going to need a bright light source.
It's amazing how quickly light attenuates with depth in water.

 

[Snarls]

Good point, especially if I use this ROV in one of my local lakes where the water is much darker than the ocean. I will be putting ~5 10W LEDs in a line on the front of the body. Not sure how much I'll need, but I figure that's a good start.

 

[PsyBorg]

the 75 mhz crystals transmitters will work through the water, for un-tethered controls.
https://www.rc-submarine.com I think sell equipment just for subs. This guy is amazingly approachable and can help you do it on the cheap. the trick will be to get the ballast to put it just at the surface, so the failsafe is to surface.

i have a couple posts on my LPL flying submarine project about setting the sub part of the project up.

keep it up, are you going to 3d print the shell?

DAM YOU Foamy!!! I went to that site last night and watched every one of them videos he has posted on the subs he built. Fell asleep watching the last one and woke up to that dudes giant head on my TV.

 

[Snarls]

Going to be all week I spend printing this thing out. Monday I printed the tail section as a test. Took 4 hours (vs 2 predicted). Yesterday I printed the rear half, and that took almost 10 hours (vs 5.5 predicted). And today I printed the top section of the front half, which took 7.5 hours (5.5 predicted). Tomorrow is the largest piece, the rest of the front half. Its predicted 8 hours, but if you've learned anything from the past few sentences, that means it will really take like 12 hours. Then after that I have to print the hatches and TPU seals. Quite possibly the largest project I've made with my 3D printer, but really it's doing all the work and I'm waiting.

 

[Snarls]

Had quite a fail today. Print of the largest body piece was looking good today. Then I went outside for an hour and came back in to this mess.

Apparently the print shifted on the X axis somehow, and quite a bit. What you don't see is that my cooling fan got caught on one of the binder clips holding the glass to the heated bed. That ripped off the cooling fan from the extruder, so I returned to see this massive layer shift plus my cooling fan being dragged around. Hurts that this was 3 hours in, but on a possible 12 hour print it could have been worse.

Already fixed the fans enough to print for the rest of the week. One of the fan blades on the main extruder cooling fan broke, so I will have to replace it. For now it is running, but a little louder and with some vibration.

I am printing the hatch doors I was planning to print tomorrow for the rest of the night, and tomorrow I will retry the large body piece. Then the weekend will be a lot of sanding, fitting, and epoxying.

 

[ElectriSean]

Gotta love 3D printing fails, it's always such a pretty mess Any idea what caused the shift? I've never seen one quite so extreme.

 

[Snarls]

Gotta love 3D printing fails, it's always such a pretty mess Any idea what caused the shift? I've never seen one quite so extreme.

Not entirely sure. I thought it may have been from the hot end getting hung up on one of the printed overhangs that tend to curl up. It seems a bit extreme of a shift for that though. The more that I think about it, the more it seems like the cooling fan nozzle got stuck on a binder clip and that pulled the x-axis to a stop before ripping the cooling fan off. Not sure how that could have happened though because the first dozen layers were fine and the perimeter does not expand as the print goes on.

 

[CarolineTyler]

Is it a requirement for your 3D prints to be watertight? I have noticed that my seemingly solid 3D prints sort of 'sweat' water will slowly make its way through the minute gaps between the layers. You may need to coat the outside with some sort of epoxy to seal them.