I'm Going Under (water)

Snarls

Gravity Tester
Mentor
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.

No they probably won't be 100% watertight off the printer. The body is printed in multiple pieces so I will have to use something to seal the seams anyway. I may use a thin layer of epoxy to coat the thing, or go with primer/sealer.
 

Snarls

Gravity Tester
Mentor
The layer shift almost got me again. The print was looking good two hours in. I went to walk the dog and when I came back another layer shift had happened. This one to the same degree as before, except my cooling fan was perfectly intact. So I'm thinking the overhangs on this print were curling up and that was snagging the hot end at some point. Luckily this shift happened at the end of a layer and I caught it early in the next layer. I used my signature high tech fix option and paused the print, forced the x axis back into position (I think I still ended up 1mm off), and started it again. The rest of the print (10 hours total) went perfectly fine.

Now all the body parts are printed. I need to do some sanding and then glue the body pieces together.The prints altogether already weigh 550 grams, but I know I'll need a lot more than that to be neutrally buoyant.

IMG_20180616_134443.jpg
Parts.jpg
 

Snarls

Gravity Tester
Mentor
This weekend was busy getting the body assembled. A good amount of sanding was required to fit the pieces together. I put tabs on the models to help make fitting easier, but I should have added a little more tolerance knowing I was going to print these with a large (0.3mm) layer height (more imprecise). Regardless, some sanding and the pieces fit together nicely. The rear section has a little warping, but it is workable.

I was about ready to glue everything together when I realized I forgot to install the nuts for the hatches and tether mounting point. For the front hatch, there are already holes for the nuts to fit. For the rear hatch and tether mounting point, I quickly drew up some parts to hold the nuts in place. I could then glue these to the inside of the body, thus locking the nut in place and sealing the bolt hole from the inside. For the front hatch I also printed a cover to seal up the nuts and bolt holes.

IMG_20180617_001739.jpg
IMG_20180617_002418.jpg
IMG_20180617_001718.jpg

For the tether, I realized I should add the receiving connector while I still have good access. I decided to use 8 female pin headers (like servo connectors). Four in a row and two rows. This is not the best choice for a connector, but it is all I could come up with to have 8 connections without going out and sourcing something special. I mounted the connectors flush to the outside of the body. Finally I put a liberal amount of silicone around the connector and all the nut holders I installed. This stuff seems great for waterproofing and for once I don't have to worry about weight. In fact more weight is currently good.

IMG_20180617_153743.jpg
IMG_20180617_153732.jpg

And finally I am ready to start gluing things together!

IMG_20180617_155341.jpg
 

Hai-Lee

Old and Bold RC PILOT
You have not stated the maximum depth and hence pressure requirement for this submersible but if water ingress does become a problem or even failure due to implosive forces there is a simple method to solve both issues.

Fill the whole assembly with a clear, inert, oil, (transformer oil). The oil as a liquid cannot be compressed and without a pressure differential there is no force trying to push water past the seals and through imperfections into the internal area where it can effect the electronics.

Just a thought!

Have fun!
 

PsyBorg

Wake up! Time to fly!
I wonder since weight is not so much an issue if you coat the entire outside of this once its set and tested with Flex Seal if that wouldn't extend its longevity and survive-ability.

I don't see this reaching any "Crush Depths" as it is after all a tethered unit built under budget constraints. But it is a valid thought for down the road if the initial unit is a success I would imagine.

EDIT: Just went back and looked at the pictures better. That's some nice quality work Snarls. That printer does well by you. As thick as those parts are I will be surprised if there is any pressure issues for a fair amount of depth.
 

Snarls

Gravity Tester
Mentor
It's like you guys are spying on me or something. I was just reading about how the industrial ROVs use cylinders with the electronics imersed in mineral oil. I am going to use a 50 ft tether so that will be it's maximum depth. Although, I will likely mostly use it in depths of 25 ft or above. At 50 ft (~15.2 m) the pressure is approximately 150 kPa (~22 psi). I printed my parts with a 5 mm thickness, with three perimeters and 25% infill. Part of the reason I did not go more into pressure and depth is because I find it very difficult to calculate how much pressure a PLA 3D printed part can take. Like you mention Bill, I just thought I'm not going too deep, and decided to use a relatively thick wall as a best guess.

Thanks for the printing complements Bill. This Folgertech printer of mine has really done well over the years. A few issues here and there, but it has always been able to make my projects a reality. Oh and you mention Flex Seal... I had the same thought and just earlier today sprayed some flex seal on the inside of the body. I don't think it would look too good on the outside, but if water is leaking in I just might.
 

Snarls

Gravity Tester
Mentor
Got the body glued up, gaps filled, and paint primed. Here is what it is looking like now:

Primer1.jpg
Primer2.jpg
Primer3.jpg


I also forgot to talk about the electronics and programming. There will be three motors, and correspondingly three ESCs. I am using three of the 12a Rotorgeeks ESCs from my very first miniquad! I set the motor direction in BLheli to bidirectional so I can spin the thrusters in forwards or reverse.

Using an Arduino Uno, I was able to set up a Wii nunchuk as the controller for the three motors. The nunchuk is simply an i2c device. Inside it has an analog joystick, two buttons, and an accelerometer. I have it set up using the joystick to control the rear thrusters, and the two buttons will control the central thruster forward or reverse. Maybe in the future I will implement some control based on the accelerometer.

Before I linked the nunchuk input to the motors as output, I first tested on LEDs. The brightness of the LEDs corresponding to the speed of the motor. This way I could test reading the nunchuk input and interpretting it before messing around with the ESCs and motors. After that, I switched to controlling the actual motors, and was able to get a satisfactory level of control. I won't be able to tell if I should change the throttle endpoints or sensitivity until I get this in the water.

PrototypeCircuit.jpg
 

PsyBorg

Wake up! Time to fly!
Thats a lot of good progress Snarls. I flashed back to lab practicals seeing the bread board and the wires. Spaghetti lab FTW!!

Just need to know why you went with army olive drab instead of NAVY colors!! We all know Army can not swim which is why they are in the army to begin with hehe. That be all you can be video where they pop up outta the water was filmed in a kiddie pool.
 

Try-n-glide

Active member
I assume you already have a go pro or similar, but if not I used a Cam Park ACT68 on a recent scuba trip. $33 on Amazon Prime, 1080 vid, and comes with a waterproof case which I used down to 80ft with no issues.

I was told by an experienced underwater photographer that I met on the trip that the camera would work better with a red shift filter (basically a piece of red plastic film placed on the case over the lense) since low energy red wavelengths are blocked first the deeper you go. 10-15 ft probably not needed. 30ft + you get a distinct blue off color in your pics.


Cool project! Keep us updated.
 

Snarls

Gravity Tester
Mentor
Thanks guys. That's just grey primer paint Bill, the final color will probably be classic ROV yellow. Good tip Try-n-glide. I have some colored gels from an old project that I can test.
 

LitterBug

Techno Nut
Moderator
Are the .stl files available for this somewhere, or is this all your own creation so far? Awesome progress!

Cheers!
LitterBug
 

PsyBorg

Wake up! Time to fly!
Bumping this back to the top.. Any more progress?

I was just pushin yer buttons over the paint job. Still looks green to me but hey don't trust my eyes for anything they see. Lets see this this swim!!!
 

Snarls

Gravity Tester
Mentor
Sorry I was busy trying to get the ROV built before leaving for the beach this past Friday, and did not put the time in to update this thread.


The wiring for the control the system is pictured here. I am using an Arduino nano as the microcontroller. To have the Wii Nunchuk and video monitor detachable, I am using balance connectors. I also have a switch to reset the Arduino in case any wrong behavior occurs and I need to stop the motors. To monitor the behavior of the motors, I am using 3 LEDs whos brightness will correspond to the motor speed. If the LEDs don't match my control input, then I know something is wrong and I can reset the Arduino.
ControlWiring.jpg


Here we are with everything wired up to the back of the circuit board the Arduino is mounted on. The tether is just a 50 ft ethernet cable. It's a cost effective method to get 6 wires in a nicely protected bundle.
TetherWiring.jpg


TetherBox.jpg


Here are all the control electronics in the housing I made for them. On the bottom are the reset button, LEDs, and balance connectors. The top has a lid that still gives access to the Arduino usb for programming. The housing is round because it is made to fit inside a spool that used to hold 3D printing filament. Now that spool is perfect for winding the tether on.

TetherPins.jpg


The other end of the tether connects to six pins. These are the pins that will connect to the ROV body. It was a tight solder job, but actually was easier than expected.

TetherEnd.jpg


And here are the pins with the mounting clip attached. Bolts will go through the holes on either side and bolt on to the ROV body. I also put silicone as a relief at the point where the tether exits the plastic. I also printed a TPU gasket to go between this plastic and the body.


I'll have to get some more pictures of the painted body with motors installed, as well as my tether spool. Right now it is going to be several days of leak testing (already found some!) and patching.
 

PsyBorg

Wake up! Time to fly!
Hey mate..

Word of advise.. tie a knot in that wire inside your cover. as that cable ages it will get brittle and more then likely break at that pinch point I assume is there to hold it. Specially if you are going to rely on that for emergency pull ups or extra oomph if it gets hung up on something. I have done tethered projects in the past where the wire casing ages and became problematic or just plain broke.
 

Snarls

Gravity Tester
Mentor
Good point, I probably should have done that. Here's why it probably won't be that big an issue. The tether is wrapped around a 3D printer filament spool, and the end that goes into that housing is the end of the tether, meaning there are many wraps around the spool that take up the load before it reaches the end going into the housing. The problem would be if I used the full 50 ft and all the load was on the end of the tether.

Ill post some pics later.