I don't know too much about the Arduino, but have seen it used alot on MAKE and hackaday.
I am planning on building an ROV underwater explorer. I am planning on having a Cat-5 tether, and using one of the wires as an antenna signal for a RC control (non-arduino at this point). With only two twisted pairs, the IP Cat-5 throughput would be at 10 mbps.
I was thinking about using an Arduino to get information from several sensors, and interface with some servos to control a robotic arm/claw.
Here is some of the functionality that I'd like the arduino to interface (and some examples of their usage):
Internal water sensor
this is probably an easy circuit, to check for a closed circuit. not really concerned about this one. water detection circuits/devices are abundant.
Power monitoring
I'd obviously like to monitor the onboard battery life.
Web interface and camera
I've heard about ethernet hookups for Arduinos.
I'm not sure exactly how the Arduino will work with a web interface. Probably, I'll have a web server running at the other end of cat-5 network, gathering and sending data to the arduino.
Servo manipulation
Anyway, anyone have recommendations or helpful links they can point me towards? Which flavor of Arduino would work best for all of these applications together?
most of what you want can be done with arduino, speed may be a factor, video is outright no though the mcu
you can collect sensor data and send it serially back to your host, same with control information from the host to the remote, that leaves plenty of wires to send a basic video signal up the wire to a certain length of cable
As you progress with your underwater ROV you may want to take note that the Freescale sensor suggested as a depth gauge is designed for dry air. The quote below is from the MPX4250A datasheet:
The MPX4250A series pressure sensor operating
characteristics and internal reliability and qualification tests
are based on use of dry air as the pressure media. Media,
other than dry air, may have adverse effects on sensor
performance and long-term reliability.
Accurate sensors appropriate for liquids are generally quite expensive. A low cost option may be to look for a surplus engine oil pressure sensor.
If you plan to use cat5, use cat5 with stranded core wires, not solid core; the flexing otherwise over time will break the wires. Stranded core will last longer, but even it will fail given enough time. As far as video is concerned, you can get interface baluns to impedance match 75-ohm coax (with composite video) with cat5 (it will only take up another pair). Another possibility (if you have the power capability at the ROV end) is to use the Arduino as a motor/servo/sensor interface, and connect it to a small networked single-board computer (like a BeagleBoard or similar), hook the ethernet pairs up, and a web/ip cam, and use ethernet for everything; controls, sensors, camera, the whole nine yards. There's probably enough pairs left over to set up power-over-ethernet (though I am not sure what the current limits would be).
What you need to make your project work depends on its mission and how mild or wild you want to make it. I think you can probably send video and rs232 communications over maybe 100' of cat5 unaided. The USB interface on the arduino is not possible over long distances unaided. Maybe the first thing to do would be to get some cat5 wire and start experimenting. You might be able to get by with a single coax wire.
I think you can probably send video and rs232 communications over maybe 100' of cat5 unaided
oh yea, that hack a day link I pointed out was doing ntsc upto around 70 feet without external amplification, doing the same without the color information I bet one could do 2x that length
rs232 can go a silly amount over cat5, course were probably dealing with ttl serial so you may not get hundreds of feet, but really the video signal is going to be the main limit of cable length
oh I have no idea, I didnt actually bother to read anything aside from the summery
it does sound low to me also, but as I dont really fiddle with long range stuff I dont have a real opinion, I do think that without proper requirements met that the video will crap out before data
then again I have never done "long range" system with TTL serial, its always been RS232 or better, and that junk probably could make it 70 feet on stretched out coat hangers
that junk probably could make it 70 feet on stretched out coat hangers
When I was kid, me and friend stretched about 50 feet of various cable - phone cable and speaker wire, mainly - from one side of his house to his room so he could have cable TV (we didn't have any coax). We ended up stapling and taping it up along walls, ceilings, and floors; it worked great for several months until his mom got a real installer to come out and hook him up right.
I've got some inexpensive NTSC color security cams that come with 60' of video/power wire. I've also run some other cams on cat3 phone wire. Baluns supposedly can get really impressive distances. The below has a possible setup to get some extra distance on an rs232 setup.
What about using a wireless RF video system (wireless surveillance systems come to mind and aren't that expensive anymore)? If you use one conductor of the cat-5 connected to the antennas on either end, this might work to give you a longer reach since all you'd be doing is providing a simple path for a signal already modulated to go through the air. It would also save you a wire for something else.
If you use one conductor of the cat-5 connected to the antennas on either end
The reason why cat5 (and coax) consist of two conductors (or pairs) is that one conductor acts as a "shield" to prevent signals from leaking out or in. While you might be able to get away with using a single conductor for video (whether as an RF transmission or otherwise), it might bleed out and cause interference on other lines (say ones carrying data); or the signal fluctuations in those other lines could bleed into the video (making it fuzzy, or even worthless as a signal) - these fluctuation could be especially true if the lines are carrying power signals for motors and such, where the brush noise would cause snow and other stuff to bleed in. That isn't to say a shielding conductor is perfect; its only as good as its ground, and even then it may still allow leakage in both directions - but it should be much less than without using it.
Ethernet cable isn't designed to work under water! At the very least I'd expect severe degradation of frequency response of long runs, and if salt water I predict a very short useful life... Choose a marine-rated cable.
The reason why cat5 (and coax) consist of two conductors (or pairs) is that one conductor acts as a "shield" to prevent signals from leaking out or in.
Coax has a shield, in cat5 there is no shield, it picks up lots of interference. However the symmetry of the pair means that its all common-mode interference and the signal is detected by a differential amplifier which is insensitive to common-mode signals. cat5 definitely leaks a lot of signal out, try placing an AM radio near it!
Ethernet cable isn't designed to work under water! At the very least I'd expect severe degradation of frequency response of long runs, and if salt water I predict a very short useful life... Choose a marine-rated cable.
I thought the outer insulation was PVC? Of course, in a chlorinated or salt water environment it won't last, and ultimately it won't last in the long run period, but I suspect (though the OP didn't give any details) that this project won't be for commercial research use but rather for fun or an ROV contest of some sort. As long as the OP makes sure to seal up the ends well, he should be fine.
Your recommendation is spot-on though if he is planning something more robust; in that case marine-rated cable and connectors are a must (just be prepared for $$$ it will cost - jeez, some of those connectors cost more than my PC! But then again, I am not dropping my PC into the ocean).
However the symmetry of the pair means that its all common-mode interference and the signal is detected by a differential amplifier which is insensitive to common-mode signals. cat5 definitely leaks a lot of signal out, try placing an AM radio near it!
Huh - here I thought all this time that the extra wire acted as some kind of (poor?) shield, but your explanation makes more sense. That also explains why you don't want to kink/stretch a cable as you run it (for ethernet).
As I have said before, I learn something new here every day!
They do sell shielded heavy-duty cat5. I've seen it used for connecting Ethernet between operation booths at concerts. This stuff is about 3/8" in total diameter and is designed to be stepped on, run over, run alongside 3-phase electrical feeder lines, etc. I can't imagine it's all that cheap, but it could last in the long run.
Another idea would be to just make your own cable snake, use flexable PVC tubing like the stuff used for fish tank filtration systems. Get a 100' reel of it and snake your cables down into it. You could run regular cat5 down this tube, seal both ends with some silicon, and that should protect it from pressure, moisture and abrasion. If you need shielding for interference, with some care you could possibly wrap the cat5 in aluminum foil before putting it into the tube.
Another nice thing about the tube is that you could just run your own conductors down it. Run a reasonable sized power cable down it to power your motors, lights, camera, etc. and also power your logical circuits with a resistor. Then run a thin shielded coax down it for the camera, and also use cat5 or phone wire for control. This tubing usually goes around $0.50 a foot, so a reel might cost a little more, but it would do you better in the long run with the ability to run the right cables rather than making it work with the easy method. Since these tubes are used with saltwater tanks, it should be able to hold up to the saltwater and anything else.
I've been thinking about this same idea, I think that once my CNC machine's working, I'll take this on as a nice summer project. There's sunken boats all over the lake my boat is on, even some sunken boxcars from a train derailment over 100 years ago, that would be neat to get a closer look at.
Everything you noted is a good idea, with the exception of something you don't tend to think about (unless you're a constant tinkerer/thinker like me) until you get your ROV in the water:
The weight of the cable.
The more you add to make it robust, the more it weighs. You have to pull that weight around and/or support it with floats. If you add floats, you add drag in the water, so you need more power. Basically you are fighting this weird power vs. weight vs. drag vs. rubustness feedback cycle. For a small ROV (something mainly meant to go into a pool), you can't afford anything too heavy, too thick, or with too much drag. You also have to keep/take in mind the weight to reel in, and carry, all of that cabling (which is why you see on the big ocean-going ROVs huge engine/motor-operated cable reels). Adding floats makes everything bulkier (and more likely to tangle).
Viscious circle, unfortunately. Your idea of putting the cable into aquarium tubing is probably the best idea; I think I would use the more flexible silicone tubing rather than the PVC, though. Perhaps even get the Cat5 in silicone clad outer sheath as well. The lighter and more flexible you can make it, the better overall, at least for these smaller ROVs.
You could also use a tube with a little extra space inside and seal the ends around the wires with hot glue or silicon sealant. The deeper the ROV goes, the more pressure the tube has to withstand. As long as the air pressure in the tube doesn't escape it, the whole length of the tube should act as a float. By varying the pressure in the tube, (say with a servo connected to a big syringe) you might even be able to adjust it's buoyancy. Since the flotation would be from inside the tube, there's nothing to increase drag aside from the tube itself.
cat5 cable could work if you're not afraid of interference, bad picture quality, short ranges, and possibly not having enough power, but if you can use the right cable for each task, it shouldn't cost too much more, but would make better in the long run.
