You can do all of the above operations on Windows without Arduino, can't you?
Where is the Arduino involved here?

He wants to use the Arduino as a USB pendrive with a software payload on it. You're better off checking out the USB rubberducky (http://www.usbrubberducky.com/wiki/doku.php?id=start) or the USB Switchblade (http://hak5.org/usb-switchblade). Arduinos are useful when you need to make your computer interact with other physical objects, its not really suited for delivering software payloads.

The hull is made out of PVC, which people on the HomeBuiltROV site have taken down way farther than 150ft so that should be fine.

I'm sure the main components of the hull can be made waterproof, but don't you need openings in it to connect external drive/control hardware to the internal controller? How on earth do you seal those well enough to withstand 150 ft worth of pressure?

We're going to try this method:
http://homebuiltrovs.com/howtosealingwireexits.html

With that much cable, you will want to make it neutrally buoyant or (preferably) have slight positive buoyancy otherwise it will turn into one big anchor.

Waterproofing at that depth would be my main concern - both for the comms line, and for the on-board electrical and mechanical systems. How on earth are you going to do it?

The hull is made out of PVC, which people on the HomeBuiltROV site have taken down way farther than 150ft so that should be fine. The tether *should* be waterproof, but we plan on snaking it through a flexible hose (made out of vinyl I think). With both ends of the hose sealed, the air will keep it buoyant.

We're currently at the "will this thing even move" stage. Its powered by 4 1250GPH bilge pumps (2 for elevation, 2 for forward/reverse). We have both 35mm props, and larger ~4" airplane props to test it with. I got a little carried away with the frame, 2.5' by 1' by 1' didn't look that big on paper. The frame is made out of 1" PVC, so there shouldn't be too much drag, but thats still a lot of ROV to be moved by two motors. We may end up cutting the thing in half. We're going to mount the thrusters today or tomorrow, and hopefully test it next week (with direct control, no arduino or tether yet).

@Hogan,
What are your experiences with the RF22 ? Looks like the RFM12B module is cheaper than the the RF24.
There is high-power version as well 500mW RFM12BP and according to EU regulations I could find, it is
allowed @ at 500mW between 869,400 and 869,650MHz. Not sure though is the 250KHz bandwidth is enough.

Thanks for the info links.

@PeterH
I feel the wifi range would be to short. Most wifi's claim 300m ... but at that distance there are a lot
of lost packets. If I could get a decent reliable signal up to 500m I'd be very happy. Hence the <1000m on paper :-)
We all know everything looks good on paper ...
 

For good, but still cheap, long range wifi, try Ubiquiti (www.ubnt.com) Their omnidirectional PicoStations are small, and the spec sheet says 500m range outside (clear Line Of Sight). If the pico doesn't cut it, you can try a nano, or rocket with an omnidirectional antenna on it. If you can deal with a directional antenna, they have some radios that have been used in 50 mile point to point setups (mounted really high up to clear the fresnel zone, and with 100% LOS obviously)

They also offer most of their radios in both 2.4GHz, and 5ghz, with some others at 900MHz.

Have you tried looking at the wireless xbox 360 controllers and the wireless PC adapter? I used to use one to connect my wireless xbox 360 headset to my PC.

Yes to the second question; in fact, if you can instead use a 4 line cable (cat-3 - used for telephones, usually), you'll end up with a lighter weight solution (less for the robot to pull around in the water, less drag means longer run time).

You should also look for (whatever cable you use) something with stranded conductors, rather than solid conductors. Note that such cable will be more expensive than solid-core conductors, but since this cable is going to bend and twist and such, you'll want the flexibility so it doesn't break. If you can find such cable that is designed for pulling, it will also have embedded in it a strong piece of monofilament or similar that can be pulled on (because it is designed for pulling thru spaces, walls and conduits in its normal usage); this will be useful if you need to "rescue" your ROV by reeling it in via the cable.

Also - if you can find it (and afford it), try to get the cable with a silicone outer jacket (and if possible, silicone insulation on the core wires); most cabling only has PVC outer jackets; the problem here is that after you use it enough, it will be twisted and flexed so much that the jacket/insulation will crack, and water will seep inside the cable, eventually running down into and inside your ROV, weighing it down (and possibly shorting things out). Plus, depending on where you use the ROV (fresh vs. sea water), the water and such will break down the PVC over time, causing it to wear even quicker.

Finally - there's always the issue of cable management; with 100 to 150 feet of cable, this isn't an easy issue. You need to be able to reel and unreel it, without it getting tangled or in the way. On professional rigs, this is done with a large spool of cable. You can do something similar - see the following page:

http://www.homebuiltrovs.com/slipringreel.html

Note - check out the rest of that site, too; it's really worth it for anybody planning on building an ROV:

http://www.homebuiltrovs.com/

So you're saying that we could also send the control data over serial to limit the total number of wires down to 4 (I assume you use one twisted pair for TX and one for RX, do you ground the other two wires?)? I was wasn't sure if there would be too much of a delay with that kind of processing. We were also planning on attaching a thin braided steel cable to it for the deeper dives, just to make sure we could bring it back up. Would that add too much weight to the tether? We will be using Cat5 intended for pulling through walls, so it should have the micro-filament inside it (I haven't had a chance to look at the stuff yet, supposedly it will be provided to us by the school). I never even thought of using that to pull the ROV back. Do you just strip off the wires at the end so you have enough of the string to tie off to something inside the ROV?

We are planning on using stranded copper, and I saw the method for waterproofing it on homebuiltROVs (where you strip each conductor and seal with epoxy). I saw the reel on homebuiltROVs too, but we probably wont get that elaborate since we're looking at less than 150ft of cable to start with. Although, this may turn into an extended project.

Thanks for the help!

Hello,
First, a little background on the project. I'm part of a team making an underwater ROV for a club project. The ROV has 4 motors that will be wired to 8 relays using this method: http://homebuiltrovs.com/howtorelays.html (except we will be using an Arduino with transistors to operate the relays). We plan to use an old PlayStation 1 controller with the PSX library (in the playground here) for control. The max depth of this ROV wont exceed 150ft, and will most likely not be below 100ft. We can get a substantial amount of cat5 cable from the school, so we plan on using that for the tether. The ROV will be powered by on board batteries, with the PlayStation controller connected to the tether on the surface.

First question, with potentially 150ft of cable, will the PSX library still work? There is going to be a substantial voltage drop over that distance, right? Would we be better off having the Arduino on the surface, and using the cat5 to connect the digital pins to the transistors?

Second question, We would also like to be able to control lighting on the ROV, and read a few different sensors (various water level sensors, temperature sensors, an accelerometer, and maybe pressure sensors). Would it be possible to have a second Arduino on the surface that communicates with the one on the ROV via the tx/rx pins? That way we could output the data from the sensors to an LCD, or to a computer.

You might be better off starting with something like this:

http://www.amazon.com/Myvu-Personal-Viewer-Universal-MA-0483/dp/B000VKY0J6/ref=sr_1_41?ie=UTF8&qid=1326742069&sr=8-41

and modifying it. I have seen cheap Chinese made versions of those sell for less than $50 USD.

i wanted to make something that would recognise if someone was walking in or out of a room. it was just two lasers on ether side of the door way, pointing at two photocells. when someone would walk in one beam would be blocked then the other. depending on which one was blocked first you could tell if they were going in or out

This is how the DOT road sensors work (although they dont use lasers). I'd go with this plan, its simple and effective.

Use a computer running openCV to do the video processing. It wouldn't need to be anything special, just a microATX motherboard/CPU combo. You could have multiple tuners with a CCTV-like setup to view each room. Connect the arduino to the computer, and use it to interface with the home automation hardware (to turn lights off an on).

With OpenCV you can set boundries. For instance, you can have it set so anything that is shorter than 3 feet (pets) wont make the lights turn on.

Wow, that's quite alarming. I imagine that at those speeds accidentally dropping into first gear could be catastrophic at any time, and plenty of other gear changes could send you off the road if they occurred at the wrong time.

In all honesty, a manually operated switch system that you're describing strikes me as so dangerous that I wouldn't be prepared to risk driving it at those speeds on public roads.

I believe it could be done safely, if you put in sufficient checks and fail safes. But without them, it's a frightening prospect.

Whether it could be done legally is another matter.

Here in the UK, it's a legal requirement to have motor insurance, and the driver has a legal obligation to inform the insurers of any modifications that could reasonably affect their risk. The list of mods on my car runs to several pages and includes some quite substantial performance upgrades. I needed to discuss some of them at length before the underwriters would agree to cover the car.

I don't think I'd have a cat in hell's chance of getting my underwriters to accept a car with a cobbled-together automatic gear shift controller without extensive and convincing evidence that it was actually safe. It's clearly safety-critical and has some very nasty potential failure modes.

How confident are you that it's safe and legal for you to drive it with your existing manually operated switch system?

The worst that could happen would be catastrophic failure of the transmission, resulting in the vehicle coasting to a stop. There is no way this could cause loss of control of the vehicle. Also, the speed is really irrelavant. Once you get up past ~40MPH you're not going to be doing any more shifting, especially on an automatic transmission.

Hello,

I have a pocket door, something that looks like this: http://www.renovation-headquarters.com/images4/pocket%20door%201.jpg.

I would like to open and close the door with the push of a button (and eventually cat + RFID tag). At first, I thought of a servo system, but I'm wondering if it is the right approach?

Could it be a magnet? When the door has to be opened, no power; when the door has to be closed, a lot of power then gradually less, only to keep it closed -- but it would have to be very strong to close the door when it is fully open.

What would be the best approach? The drywall is already done and I'd like something "external" if possible.

Thanks!

For a magnet, you can get electromagnets that are designed to hold doors open. You could use springs to make "open" the normal state for the door, and have the magnet hold it closed. You could also add some sort of hook on a servo, so that you could turn off the power to the magnet and still keep the door closed.

Your best bet would probably be an actuator, or some sort of string and pulley system, mounted to the top of the door. If you look around on instructibles.com and youtube, a few people have made automatic sliding doors that mimic the ones from Star Trek, you could probably get some ideas from those.

Hello,
When you upload a new sketch file, it will overwrite the old one, so no delete is necessary.

Yes, all those terrorist websites like Ebay, Amazon, hobbytron, and Estes. Just use the commercial fireworks/rocket igniters. They're cheap, safe, and legal.

At $2+ each plus postage, maybe not that cheap.

If its for a movie he's not going to need a lot of them. $2+ postage is a lot cheaper than a visit to the emergency room when he lights his hand on fire trying to make something out of matches and lighter fluid.

If you need a electric spark, the piezo spark from disposable fireplace lighters or disposable electric cigarete lighters might be an option. if you need an ignitor, any good terriorist site has various methods.

Yes, all those terrorist websites like Ebay, Amazon, hobbytron, and Estes. Just use the commercial fireworks/rocket igniters. They're cheap, safe, and legal. If you really need a "spark" and not a "burn" you can use a disposable camera flash circuit to make a very large spark.

I would advise against using any kind of flammable vapor (gasoline, alcohol, ethanol). You're not going to be able to judge the size of the fire. If it happens to be particularly warm out, you could end up with a sizeable cloud of explosive vapor. Theres lots of examples on Instructables.com of people making arduino controlled propane fireballs for shows.

I did not see the bit about using gasoline when I first replied. Actually, I would advise against attempting this project at all. If you have to ask how to ignite it, then you don't have the experience required to do it safely, especially when other people may be near.