Has anyone used wired comms (not wireless) to their ArduRover project? I'm looking at how I might use ArduRover for an underwater ROV and the first requirement will be that I communicate via a wire (preferably a single twisted pair). The idea would be to use the Mission Planner to navigate the ROV on the water's surface to the desired location and then go into manual mode and dive underwater.
Another issue I'll also have to tackle will be how to modify the code for vertical thrust (to move the ROV up and down).
I think you would be better off starting with the Copter code and re-working the AP_Motors library rather than starting with Rover. Vertical thrust will fall out of the motors library changes. Wired comms will be no problem
My ROV project currently uses RS-485 bidirectional communication over a twisted pair between the ROV and topside control station.
You also might look at how the OpenROV project communicates using a LAN to power line converter over a twisted pair between the ROV and the topside control station.
@Thomas - RS485 is exactly what I was thinking, too. Would you mind sharing your setup? Yes, I've considered using "HomePlugs" but haven't tried them yet.
@Craig - Thanks for the heads up. Started to scan through some of the code and have to say it's over my head. Is anyone else working on modding the code for ROV use?
You know on subs, we used a towed array to for coms. It's basically an antenna with a float on it. I've been thinking about doing a submersible rover, and it's the cheapest thing I could think of. Bassically you can have 15 feet or what ever your max depth is on a clock spring reel. So as you go down the spool spring would fight the float to extend line. When it comes back up, it will wind on it's own, and it will give you a good visual indicator on the top of the water.
one last thought, a cheap tape measure guts might just work in that application.
@Ben - That's actually been done. But in my particular application the risk of snagging the wire would outweigh the benefit I think. One advantage of having a tether is that if you get lost (and in poor visibility and no sonar that happens) you can always pull on the tether to retrieve the ROV.
What do you plan to use for a tether?
Presently Blue Robotics is using Cat 5 twisted pair cable, which is around 10mm in diameter, as they have their batteries in the ROV.
One of the members on homebuiltrovs Link suggested using 1/4" Flexzilla air hose Link. This hose is extremely flexible. The Flexzilla hose is advertised as being able to lay flat right after uncoil and remain flexible even when it has been completely frozen.
In some cases you might just want only two twisted pairs if you keep the batteries in the ROV. One twisted pair for communication and one for video.
Normal Cat5e patch cord (stranded) is too weak IMO. "Tactical" grade Cat5e is considerably tougher, and yes, stiffer. Low cost, off the shelf, tether cable is pretty hard to find. You either have to buy excess stock from a manufacturer or make some compromises.
Looks like it might be pretty stiff given the thick shield braid.
A heavy tether cable just puts an extra load on the ROV.
How tether thickness and weight effect the ROV really depends on how big and heavy the ROV is. ROV tethers, in general, tend to be heavy and fairly stiff. Most I've worked with have thick, PVC covers and tightly packed centers. On my last build I used tether I bought from Outland Tech. Very nice cable, but I would consider it pretty stiff.
Standard CAT5e cable has an air void which is generally a bad idea for use underwater. Direct burial CAT5e cable uses a gel to block water ingress but is generally solid core wire so not suitable for our purposes. Tactical CAT5e (such as what Gebco offers: http://www.gepco.com/products/proav_cable/network/4pr_cat5Eheavydut...) looks to be the best option. Just be sure to get the stranded core version.
Another way of getting around this, and offering the ultimate in thin tethers, is to consider using fiber optics.
Wow! Looks awesome Victor. How far a long is the project? The schedule shows a completion sometime this month, is that likely to happen?