Let's Make Robots!


This spring I’m trying to build a ROV for me and the kids to play with in the summer. I’m trying to make it as small as possible so it’s easy to carry and bring along.
Instead of feeding power down the tether I use a 12V 1.2Ah lead acid battery as the power source and ballast. There will be a webcam onboard and the only signals going down the tether is serial signals to control the motors and video signals up to my laptop.

The control box contains an Arduino and two thumb joystick. The Arduino translates the joystick input and sends serial information down to the serial motor controller http://www.pololu.com/catalog/product/1110 . Sins I have 3 motors I had to daisy chain two boards.

The hardest part to get built is the stuffing box. Luckily my brother has contacts that could machine them for me and they came out great.
Still waiting for the USB webcam to arrive and can’t do much testing before it has been mounted. When everything is assembled and found working (and balaced) I will coat the ROW with Plasti Dip to seal everything. http://www.plastidip.com/home_solutions/Plasti_Dip


Project update: 23/5-2010

Problem 1
With all installed the ROV has a negative buoyancy.
Remedy: Buy more 40mm PP tube and extend the connection joints. Some wiring also has to be lengthened.

Problem 2
My 35mm propellers at 1000RPM is not giving me enough thrust.  So I will be ordering some new 3-blade 52mm ones and hope that they will give me some more thrust.

Problem 3
With my 15m of cable the USB webcam is not giving me an image. The pc recognizes the device but the cable (a CAT5) is probably too long and I’m running into some sync issues.
I found some USB to CAT5 extenders that might help.
I need to hack them in order to make them fit into my tube and control box.
Has any of you any experience with USB to CAT5 extenders?

-- end of update --

Project update: 10/8/2011

This project has been on the backburner for over a year. I was about to scrap the whole project but my brother convinced me to have another go at it. So now I’ve pretty much ripped the whole thing apart. Thrown away my control box and started over.

So this is my new approach;

Using a USB based webcam didn’t work. So I have ordered a cheap CCD camera of eBay, together with a 7” LCD monitor. This is probably a good think as I can mount the joystick and monitor in a bigger box and there will be no need for a PC.

I have got the new propellers and rebuilt the shafts. It looks better but has yet to be tested.

One of my main problems was how to get a waterproof connection from the tether cable through my PP plastic plumbing parts. And mounting it on a round pipe made it even worse (PP plastic is almost un-gluable). So my solution became to add a 45deg t-fitting on the motors. That gave me a flat fitting with a regular garden hose adapter. These adapters come in and variety of different materials and will be much easier to work with.

I’m still not certain that this ROV will ever work, but I have a better gut feeling about this design than the last one.
The image shows most of the parts assembled and nearly all the of the wires.

My plan is to have the video signal and power in on the left side and the serial, error, and reset on the right hand side. They will be joined into a CAT-5 cable going to the surface.


-- end of update --


Mounting the thumb joysticks   

Inside the control box

The frame


Battery in place

 The main structure is mad from 40mm PP- pipefittings

The serial motor controllers and voltage regulator

Not much room

The stuffing box

Fill up with grease

Stuffing box assembled (updated: 19-05-2010)

And finaly mounted

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

While watching an episode of "How It's Made" they actually quickly went through the production of a ROV. Their props were ducted rather than being open. First suggestion, enclose your props. Second and much more expensive, rather than working with a microcontroller, have you considered using a single board computer and then sending the video feed via a compressed data signal rather than trying to send the raw video feed?


Internet searches have come a long way in a year. The BCH4105G is indeed a USB extender controller chip.


Thank you for your input Birdmun.
About the ducting. Did they say anything about the purpose? Is it just to protect the propeller or does it give more power or control?
With my 3D printer I will probably makes some anyway, even just to make it look cool :-)

About the video, there is limited space inside my 40mm tubes so any equipment has to be really small. I have a go with my CCD setup first and see if that works over a CAT-5 cable.

The ducted fans produced a bit more concentrated thrust. Wikipedia seems to somewhat agree with my belief.

I doubt if the extender you linked to will solve your signalling troubles. They look like passive connectors to me. Or at least the description on the e-bay page fails to convince me otherwise.

A passive cable is what you've got already. Unless these products actively process the signals (repeating or boosting), they will not make a differene.

I got the USB extenders in the mail today and opened them up to have a look.
There is some sort of electronics going on inside and it’s not just extending the wire. There is one large capacitor and an assortment of surface mounted component, including an IC with the marking BCH4105G but I’m not able to find any more information on it.
I will not be able to test it further as I’m currently working on another and bigger project.
I’ll keep you posted when I get back to the ROV project.

The cap plus IC suggests (at least the possibility of) boosting signals. Powered by the USB 5V supply.

I agree.
The good thing for my project is that it seems that the system isn’t using more then 4 wires of the CAT5 cable leaving me enough  wires for motor control signals.

Agreed, I can't think of what else it could be. I couldn't find any details on the BCH4105G either, but it's probably just a high-speed buffer IC for maintaining signal integrity.

Well for that price it’s worth a try.
Hopefully they are not just passive connectors and utilize the 5V from the USB port to boost the signal between the units. One would probably not get the full bandwidth of a USB 2.0 but if I get enough for a video feed I’ll be happy.
I will give you a report when I have tested it out.

Guess you would have to make the led lamp pan/tilt-able also?