Build log for my DIY ROV


#1

Hi all,

i’ve been planning the build for a few months now and have made a start. It’s a ROV intended for undersea exploration down to 100 metres (330 feet). It’ll have a rectangular frame with 4 lateral thrusters, one on each corner, all at 45 degrees, and 4 vertical thrusters (because four motors running at 50% is more efficient than two motors running at 100%).

The frame will be made from PVC tubing (not thick-walled since the frame will be fully-flooded so there’ll be no pressure issues), the electronics housing will be an acrylic cylinder with a crush-depth of 150 metres.

On-board will be an Arduino Uno for control and a Raspberry Pi for video. The ROV will have 6 DOF so no need for a camera-tilt solution.

Topside wil be another Arduino that relays control commands to the Uno on the ROV. The ROV should be about 30x20cm (1 foot long and 8 inches wide).

The ROV will be powered by an on-board LIPO with a PDB for the ESCs and other electronics, the tether will be Cat5 with floats.

Of course, no plan survives the battlefield so there will probably be changes.

My first step was to get this PS3-controller clone :

working with this USB-host shield :

http://www.ebay.co.uk/itm/Arduino-UNO-USB-Host-Genuine-Keyes-Shield-ICSP-Funduino-Google-ADK-Flux-Workshop/121978946221?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649

to control these motors :

via these ESCs :

using the PS3USB part of this USB library :

Having had some frustration trying to control the motors with potentiometers (without them smoking!) and analogue joysticks (which were poorly-made), the solution above worked as soon as i coded a rough control sketch.

I’ll be posting updates here and no doubt asking questions! I’ll also do videos and post to YT now and then.


#2

Sounds interesting! How do you plan on transmitting video from the ROV?


#3

Video transmission will be from the Pi (as described above). Cat5 handles all of it. On-board power so no PoE.


#4

And btw i’m not one of these guys that starts off with great plans and then doesn’t post anymore, my PVC tubing arrived today (stronger than i imagined it - furniture-grade with 4mm walls). I’m going to see this through, hopefully for under £500 - i’ve spent £300 so far).

I’m determined, though my skills are patchy (a little electronics but a lot of coding experience).

I’m going to use it to explore the Celtic Shelf, just off what we call Land’s End (the SW tip of Britain, the Cornish coast). My interest is in Human origins and ancient civilisations.

Modern-day fisher-people keep finding lots of stuff (arrow-heads, sabre-tooth tiger teeth, etc …), and there’s not been a lot of exploration in the area i plan to explore.

My next step is to see what volume of space is needed for the Uno and the Pi plus their associated electronics, then i’ll know what size of acrylic cylinder is needed to house it all. It’s interesting to see what overall shape the device needs as you think and re-think what’s needed to get the job done.

Having got my hands on the PVC tubing today my motors now look small and potentially under-powered - but time will tell :slight_smile:


#5

i’m building an ROV with the same motors & 65mm props from Graupner. I’ve made ducts with <1mm clearance to improve thrust but i havent been able to test them yet. I think they should be powerful enough motors though. Keep us posted! And i’ll let you know if i get to test them anytime soon :slight_smile:


#6

This sounds very interesting. Please keep us updated.

One point many novices overlook is the need for plenty of controllable thrust perpendicular to the view of the camera, especially in open water where there is significant current. Current is never into or through a wall or floor, always parallel. If you want to hover and look at a spot on the floor you will have to fight sideways current to do so.


#7

Yes, please let me know when you’ve done some testing. The props i got are ~40mm and i think they will be too small, i’ll look on Graupner thanks for the heads-up :slight_smile:


#8

That’s a good point, hopefully 4 of these motos will be enough, i’m not sure of the currents in the area i plan to be ROVing about, circled in red in the attached depth-contour map.Depth-contour-map-of-the-Celtic-Sea-red-circle


#9

I notice that Graupner only seem to sell R versions of their 65mm props, i think the R means reverse?
I’m thinking that i need each pair of lateral thrusters (forward and rear) to rotate in opposite directions to avoid a net thrust in one direction due to centrifugal force, is that right ?


#10

The R means they’re designed to spin “right” only (counterclockwise if you look at the tip, i think), so they may be wrong for your design. They also usually come with an M4 screw mount, so bidirectional drive would require you to modify that as well.

after combining all this with the relatively high cost of bidirectional drivers i came up with a ridiculous design that uses a total of 12 thrusters (like yours they’re all at 45deg) That all just spin in one direction for full translational and rotational movement… not sure if that’ll work, though… i’ve only just finished the frame last week

Look for Doug Jacksons (from svseeker.com) video on his ROV thrusters on youtube for help in picking a propeller for your design. He’s done some serious testing. (And i think he hangs out in these forums as well)

And let us know what you ended up using :slight_smile:


#11

Thanks mail1, that’s a hell of a frame, wow! Is that just laser-cut sections or are some 3d-printed (hard to tell due to transparency). I bought one of those cheap Chinese cutters from Ebay and i’ve modified it for a larger bed etc, a 3D printer is on my list.

Admirable work, it looks like some kind of SciFi machine :slight_smile:


#12

Made a start on the frame :

Brief video showing the gamepad control :


#14

Bought a 3d printer so will be using that to make the chassis rather than pvc pipes.


#15

Looking into it, even a fully-flooded 3d-printed hull would probably deform internally at 100m, luckily i have a friend who does vacuum-forming so the hull will now be vacuum-formed.

I’ve nearly finished the thruster design, just waiting for some parts to try out, then the next step is pressure-testing the electronics enclosures, which will probably be made from polycarbonate but i may try out standard acrylic too.


#16

I think you will find strong currents at the location marked, but make use of “slack tide” for ROV operations.


#17

The “R” or “L” marked on the prop signify if it is right or left handed.

As a guide, make your right hand into a fist, with your thumb sticking out. If you accept that the prop rotates in the direction of the roll of your fingers, from the hand to the nails, then the thumb points in the direction of thrusted movement, and vice versa for the left hand.


#18

Thanks Martin.

I’ve completed the first thruster prototype, controlled by arduino uno with a USB shield for the PS3-clone gamepad, the problem is that it only uses 0.3A at full throttle (forward and reverse).

I’ve programmed the ESC with a card so that it has the right timing for the number of poles and should use as much power as it needs, the ESC and motor can use10A comfortably.

Video here :

Any ideas anyone as this has had me stuck for 2 days now.

Cheers, Gary.


#19

Some interesting reading on the Celtic sea :

http://plymsea.ac.uk/2132/1/The_oceanography_of_the_Celtic_Sea_II.__Conditions_in_the_spring_of_1950.pdf


#20

Now i feel stupid, someone pointed out that i was running the motor in air so it had very little work to do, hence the little current draw!