My ROV project

What are you working on .... Show off your Rov's Projects here.
Post Reply
User avatar
jenswilly
Posts: 12
Joined: Jan 28th, 2013, 4:34 am
Location: Copenhagen, Denmark

My ROV project

Post by jenswilly »

Hi, everybody. This is my first post here and I though I'd start out by briefly describing the plans for my ROV project.
Comments, suggestions and critique are, of course, more than welcome.

Frame and thrusters
The frame will be the typical PVC pipe "box" and for thrusters I'll be using six converted bilge pumps (shroud and impeller removed and mounted with propellers).

Each motor will be controlled by PWM output from the onboard microcontroller by way of a MOSFET driver and a low-side MOSFET switch (no H bridge since the motots will just run one way).

Control
The control system will consist of a top-side controller and an onboard controller.
The "brain" of each controller will be an LPC1347 Arm Cortex M3 microprocessor with custom firmware.

The top-side controller is responsible for transmitting inputs to the onboard controller.
Two input devices will be used: a USB joystick (a Saitek Cyborg EVO – since that's what I already have) and a 10" LCD touchscreen (this one). The USB HID inputs from joystick and touchscreen are parsed by a FTDI Vinculum-II USB host IC (also with custom firmware) which sends the parsed data to the LPC1347 which sends commands to the onboard controller by RS-485 over cat5 cable.

Video
The composite video signal from a fixed camera module will first get an on-screen information layer added by a MAX7456 controlled by the onboard MCU.
The video signal is then run through a video balun to be transmitted over cat5 twisted-pair cable to a matching video balun and then to the LCD monitor.

The video overlay shows battery voltage, current thruster output and maybe stuff like compass heading and depth (I haven't settled on the necessary sensors yet).

Power
Power will be provided by LiPo batteries both on board and at the top-side controller. 14.8 V probably.

Stuff remaining to be figured out
Obviously, lots of stuff still need to be figured out. Among these are:
  • - exactly how to pot/waterproof wire connections (Qsil 216 silicone, Hysol 3421 epoxy, Bulgin Standard Buccaneer sockets, …)
    - lighting (1 or 3 W LEDs, I suppose)
    - depth sensor
    - compass module sensor
    - accelerometer for attitude?
    - fail-safe issues
    - a cool name for the ROV :)
    - all the other stuff I haven't thought of yet...
/Jens Willy
User avatar
jenswilly
Posts: 12
Joined: Jan 28th, 2013, 4:34 am
Location: Copenhagen, Denmark

3D printed parts

Post by jenswilly »

I made a few parts on my Replicator 2 3D printer just to see how solid they would be and how well PLA handles water. The answer to both is, "Surprisingly well".

I made a motor shaft-to-M4 adapter in order to mount a propeller on the motor and a propeller shroud (I guess I could even go one step further and make a real Kort nozzle with an airfoil shape but I doubt that I would notice the improvement in efficiency).

Here's a few picture. First the models (from ViaCAD 2D/3D), then the parts (the propeller is mounted to the adapter with an M4 threaded rod) and finally the assembled thruster:

Image

Image

Image
martinw
Posts: 91
Joined: Sep 20th, 2011, 11:02 am
Location: Aberdeenshire

Re: My ROV project

Post by martinw »

Nice work on the 3D printer :)

Martin
User avatar
jenswilly
Posts: 12
Joined: Jan 28th, 2013, 4:34 am
Location: Copenhagen, Denmark

Progress

Post by jenswilly »

The electronics are coming along nicely: I've got USB joystick and touchscreen controls working, RS485 communication between the top-side and onboard controllers, compass IC,video overlay IC and depth (pressure) sensor working.
I still need to add calibration and pressure-to-depth conversion for the pressure sensor and I need to make the video overlay look pretty. But at least it is working.

And here's another 3D printed part: the tiltable camera mount so I can tilt the camera up and down.
Image
Citrushund
Posts: 14
Joined: Nov 25th, 2012, 12:27 pm

Re: My ROV project

Post by Citrushund »

Hey Jens

I like your work on your 3D printer, thumbs up for even having one :)

I am a Dane my self, but living in Germany now. My own progress is far from yours, doing most of my work in aluminum.
If you want to have a talk with another Dane about our projects, feel free to send me a pm :)

Martin Falk
User avatar
jenswilly
Posts: 12
Joined: Jan 28th, 2013, 4:34 am
Location: Copenhagen, Denmark

Progress update: power board

Post by jenswilly »

My original plan was to power the thruster motors (six of them) by driving MOSFETs with an MCP14E4 MOSFET driver – as described here.

And that worked fine. Except that running even one motor will make more noise than a bunch of kids at a Justin Bieber concert. Electrically speaking, that it. And while the LPC1347 MCU handled that just fine, the MAX7456 video overlay IC did not like it one bit and kept going into reset or simple displaying garbage characters.
Granted, a lot of noise can be removed by using capacitors and coils on the power supply and by careful layout of the PCB. And my current setup is on breadboards with about a million of jumper wires which makes for really crappy noise tolerance. But I'd still rather design something that will work on the breadboard setup because then it will certainly work on a proper PCB.

So I decided to completely separate the motor circuits from the digital components. I'm using a FOD3180 "MOSFET gate driver optocoupler". Which is an optocoupler designed to directly drive a MOSFET gate by being able to both source and sink a fairly large current (in this case about 2 amps) in order to switch the MOSFET on and off fast (fast switching is good since slow switching will result in power being dissipated as heat in the MOSFET).

I designed a "power board" PCB that contains the MOSFETs, optocouplers and flyback diodes and capacitors for all the motors. A Micro-MaTch socket will connect the board to the main On-board Controller board by a 12-wire ribbon cable.
The PWM pins on the MCU will control transistors (plain BC847 NPN ones – they switch fast enough for 20 kHz PWM) that will amplify the MCU pins' puny 4 mA to the 10 mA required by the optocoupler LED using a current limiting resistor.

Another advantage to this approach is that if anything goes up in smoke, it will most likely be the motor MOSFETs and now, if that happens, I don't need to replace the entire controller board (maybe I should even make smaller module boards – like maybe only two motors per board).

Here's the schematic for the power board: Schematic PDF

And here's a rendering of the PCB:
Image

Now off to ITead to get the PCB fabricated...

Components

- IRLS3034-7 MOSFET
- S5BC flyback diode
- FOD3180 MOSFET gate optocoupler
- 100 µF capacitor, Panasonic EEEHA1E101XP
- 100 nF decoupling capacitor (generic 0805 ceramic capacitor)
- Tyco Micro-MaTch 12 pin connectors
Post Reply