By Jared Hornsby
After planning this a while I've finally started work on an ROV (Remote Operated Vehicle) to survey beneath the water in our lake - among other places! The basic construction is from 110mm PVC soil drainage pipe, some 50mm PVC pipe, brackets, bilge pumps, misc RC kit bits, and one high resolution dome CCTV camera with Infra-red functionality. The PVC pipe is rated at a good pressure, so it's perfect for basic sub-level diving.
I've modified 3x 12v bilge pumps to act as thrusters, which will be running 70mm props. I'll run the two side thrusters through a v-tail mixer for forward-reverse-turn, and the upper thruster for depth control. Bilge pumps are great for this, full of torque, completely waterproof and most importantly - CHEAP!
I could make the two side thrusters rotate and save on a motor, but I want to keep this first version simple.
To make a drive unit: Take one Bilge Pump, remove the filter screen, then cut through the impeller housing to expose the impeller and motor housing. Prise off the impeller then simply make a boss to fit a suitable size propeller.
Illumination will be a pair of high output LED spot lamps, or a light bar - I'm somewhat undecided on that at the mo.
Video signal will be fed via a multicore tether back to the truck/shore/boat then to the monitor/laptop - including basic heading information, depth, and voltage status. The tether will also act as an extended low impedance aerial.
Power will be from two internal 4ah sealed lead acid batteries, plus a shore fed +12ve line for the spots / aux units (but not on this ROV).
It'll be driven (for the time being) via my awesome Robbe F14 setup. Then, once it's water-proven I'll move on to more complicated methods of control, add a modified Fish-Finder and a manipulator arm. The receiver has DSC capability, so I can easily either intercept the PPM signal in future versions, or wire up my F14 to use this instead of the 40MHz signal.
Should be at a testing stage in time for our next Free-Sail meeting on Sunday 20th January 2013.
The build cont'd...
I've now completed most of the build - the prop guards are on, the Skid is glued and mounted, the tether junction box is encapsulated in wax (it's better than Epoxy for sealing) - renewable for repairs/modifications. With that finished the majority of the wiring is completed.
I now have to work out how I'm going to cram 6ah of lead acid battery power into the ROV, along with 3x ESCs, a video overlay on screen display (OSD) unit, the receiver (RX) and the electronic switching for the lights.
Once I’m happy with all of that, then I can make up the E-Case as I like to call it. I'll be using a flight case to hold everything together in one tidy unit - main controls, monitor, power supply, and eventually Manipulator arm controls will all be in there - it'll make for a clean organised and transportable package.
One of the photos shows the “rats nest” of wiring which need to be sorted, including some heavy duty Ferrite suppression for the control feeds and video signal - it's awfully tight in the hull.
I'm hoping for it to be at the point for basic testing on the 20th Jan 2013, probably at half power and no video, but it'll prove the concept at least.
I even found 20mtrs of Ethernet cable for the tether in my stock which I thought I didn't have. It's not as long as I would have liked, but that’s what prototyping is for. Adopt, Adapt, Improve.
The MK2 is already on the drawing board!
...the build cont'd (3/2/12)
Due to work, Landrover and Home commitments, the ROV took a back seat
for a few weeks, but tonight I’ve managed to press on and complete the MK1!
As you can see, (more photos above have added. RC) I've made the system more portable, and stiffened up the skids (they were very flimsy with the ballast added). You may notice the Tether has bulked up slightly with an extra cable - the ROV will now be drawing power from a Shore based 12V battery temporarily during testing - eventually I'll make a caddy so the system can run two batteries at either 12V or a switched 24v system for the MK2.
I've had to go Shore based as the motors actually draw approx. 3.2 amps EACH, meaning even with two 3ah batteries on board the ROV would last less than 30 mins in the water - and that's without the camera and light working.
Annoyingly, the camera I had intended to use decided to die for some reason - it was cheap, so I'm not surprised at all. I have since purchased a more robust camera and decided to run it "static" in the
window until I'm happy everything works as it should. This itself has created a problem - the Infrared lamps reflect terribly on the inside of the dome, giving terrible lens wash. These have been unplugged for the time being, until I can fit an external IR lamp on the Starboard side outrigger.
"Normal" illumination - some ROV builders seem to like using the 50w 12v Halogen spotlight bulbs for lighting - these can draw in excess of 5amps and waste a LOT of power due to heat, so I invested in a 1.3w LED replacement unit that outputs an equivalent 20w, pure white light and draws approx..1amp, with minimal heat!
Ballast is going to prove very awkward - this thing is VERY buoyant! I need it to be neutral, but keep it just above the surface so that the vertical thruster works properly during the initial dive from release,
and the tether will need to be setup as well with foam pucks.
Just as well I don't mind getting wet, really :)
The control system:
I have hijacked the PPM signal from the Trainer port pins on the main circuit board. This has many, many advantages - the main ones being the signal is now fully shielded and has a significantly stable range (100M+), and uses less TX power due to the RF system being disabled.
Internally, the RX has a DSC port (Direct Servo Control) meaning I don't have to hack the RX to find the RF module! Once again, I'm using less power on the RX, too.
The monitor pulls under 1amp, meaning a cheap(ish) sealed lead acid battery is more than adequate for the control box power.
(17/2/13) The first test run...
Well, it didn't go too well! Water leaks which then caused to control problems due to wet electronics. I took the RX out, dried it off, shoved it in a plastic bag, and bodged a temporary shoreside seal - it slowed the water ingress enough to get the ROV to run out about 5M and get a good, stable 2M dive before the seals went and it started to ....erm..... "Auto-dive" =)
Exposure on the camera due to the water leaks made the two lenses condense, so there was nothing to see! When back in the workshop I found a small fracture around the front window and a gap around the tether entry point - both will need repairing before I can re-test.
After the damage to the receiver was narrowly avoided, and the fact it seems the motors I've used hate the speed controllers (possible poor windings on the coils) I've decided to put my electronics head on, and go back to good old fashioned 5A relays for a control system. This will enable a more robust, but "clunky" control system (off or on and forward reverse motor control).
I'll seal the front end with a pressure cap and "pot" the camera (encase it in Epoxy) inside its own capsule - I may add the IR lamp previously mentioned.
It didn’t go exactly as planned, but still, that's why I love my (many) hobbies - if it was easy, there wouldn't be a challenge!