Long distance video transfer over wire / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Long distance video transfer over wire

1 2 3 next

Move Thread

LAN_403

the_anomaly

Wed Jan 02 2019, 03:01AM

Registered Member #19 Joined: Thu Feb 02 2006, 03:19PM
Location: Jacksonville, FL
Posts: 168

I'm looking into building an underwater ROV. I plan to use an umbilical cord for power and control. It will dive up to 300 feet and be able to stray 200ish feet from the boat so I need the umbilical cord to be 360ft long -> call it 400ft. I plan to use a 1080P HD camera for a live stream. Does anyone have a suggestion for video transmission over 400ft? I was thinking ethernet but I would be a bit beyond its limit.

Also, if anyone has any comments or advice about ROVs I'm all ears. Going to use it to check out reef sites off the Atlantic coast.

Patrick

Wed Jan 02 2019, 03:25AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

our fellow 4hv'er Uzzors2k has some work you should review.

He used ethernet video and 130 volts to power it. i wonder if a flight control board could be used for station keeping under water instead of its intended flight purpose?

hen918

Wed Jan 02 2019, 11:30AM

Registered Member #11591 Joined: Wed Mar 20 2013, 08:20PM
Location: UK
Posts: 556

differential analog video over CAT 5 is probably the best approach. I've used Ethernet before and the lag is bad. Makes an ROV much harder to pilot (

I'm on the far right of the photo at the bottom).

Your biggest problem will be sealing the control box. at 300 ft you will have almost 10 bar of water pressure which will find its way into every crevice. Vacuum potting is a must, but if you want to keep things not covered in hardened goo, filling the control enclosure with pressurised oil is the way the big ROVs do it. a normal gland will start to leak at 10 m; the slightest scratch or on the cable, pipe or wire going through it will cause a leak.
The best approach seems to be connectors designed for the job on the outside of the enclosure, but these are expensive.

the_anomaly

Thu Jan 03 2019, 12:21PM

Registered Member #19 Joined: Thu Feb 02 2006, 03:19PM
Location: Jacksonville, FL
Posts: 168

Thanks guys!

I was thinking of building the body out of aluminium, I'll look up cylinder pressure math. Filling with oil sounds like a pain. Following the chart below, o-rings don't sound so bad unless I am missing something:

As for cable connection, what do you think about running a cable through pipe and then filling the pipe with epoxy (maybe 2" pipe length). The pipe attaches to the ROV with a taper thread or o-ring seal. The terminations are made inside the ROV so I would have to remove the back plate to detach before un-screwing the pipe.

hen918

Thu Jan 03 2019, 02:40PM

Registered Member #11591 Joined: Wed Mar 20 2013, 08:20PM
Location: UK
Posts: 556

the_anomaly wrote ... what do you think about running a cable through pipe and then filling the pipe with epoxy (maybe 2" pipe length). The pipe attaches to the ROV with a taper thread or o-ring seal. The terminations are made inside the ROV so I would have to remove the back plate to detach before un-screwing the pipe.

That sounds like an effective method, the epoxy inside the pipe will probably seal OK against the cable, a quick release connector (or other air fitting) can cope with 10 Bar without leaking if it's in good condition, the most probable ingress point will be the seal of the fitting onto the enclosure. When working with air systems that always seems to be where it leaks, and a few cubic inches of air escaping from a fitting over the course of 10 minutes is completely unnoticeable, but a few cubic inches of water could destroy everything.

Another often overlooked ingress point is the gap between the wires in the cable and their sheath, as well as the gaps between the strands of copper and the plastic. Water can wick down there with no external pressure at all. Both ends of the cable need to be really well sealed, otherwise a single leak anywhere can compromise the whole system.

I'm not sure what the porosity of a typical rubber cable is under several bar of pressure, but I would try to run all the cables inside air pipes which will remove the risk.

The only trouble with your method is that it's difficult to change anything; you have to desolder or remove the plugs on the things inside to remove the pipe from the fitting, and every time you do that you are increasing the risk that you have made something leak. A vacuum pump can help tell you when something has a leak, but the differential pressure at 300 ft is greater than 1 atmosphere, so a pressure vessel you could put it in would be more useful.

Patrick

Thu Jan 03 2019, 06:16PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

I was thinking you should use AN hose and fittings. the hose would be for ligiud tightness and strength if (when) you get a moderate snag on the rocks or kelp. run the wires through the center of the smallest diameter they make. cant remember which one is smallest though. in the aviation buisness ive never seen any of our work smaller than 1/4 inch (#2 size) Im just not sure the flexibility would be good, or if 100 ft of heavy hose would cause your bot to be un-controllable or just sink straight down.

2Spoons

Thu Jan 03 2019, 10:26PM

Registered Member #2939 Joined: Fri Jun 25 2010, 04:25AM
Location:
Posts: 615

Like this?

I know the guys who built this. They use optic fiber, and can do 4k video to 1000ft.
As I recall they started out using co-ax for the video (at 1080p), and the first body was built in an aluminium dive tank.
Cat5e or Cat6 ethernet cable is going to be a problem under pressure - too much empty space inside the cable means it will crush, and its impedance will change in unpredictable ways.

Patrick

Fri Jan 04 2019, 07:33AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

i think he's trying to avoid the use of batteries.

...

Fri Jan 04 2019, 08:26AM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

You could try one of these 'hdmi extenders'

They appear to upconvert the video to some RF frequency and transmit it down a differential pair. They are spec for 1km of phone cable at 1080p.

You could probably also get away with 100mbit ethernet at 400ft if you use cat5e or cat6 cable. As long as you are running full duplex (ie, no hubs, just a direct link from nic to nic or nic to switch etc) there is no chance of collision and the whole issue of 100m time of flight limitation for collision detection is not an issue, in fact you can use a 'dumb' converter to optical and go for 5km with normal 100mbit full duplex. It is just a matter of ensuring that there is no too much attenuation/crosstalk in the cable, and people report even 200m of cat5e will work with 100mbit full duplex.

2Spoons

Sat Jan 05 2019, 09:49PM

Registered Member #2939 Joined: Fri Jun 25 2010, 04:25AM
Location:
Posts: 615

the_anomaly wrote ...Thanks guys! I:As for cable connection, what do you think about running a cable through pipe and then filling the pipe with epoxy (maybe 2" pipe length). The pipe attaches to the ROV with a taper thread or o-ring seal. The terminations are made inside the ROV so I would have to remove the back plate to detach before un-screwing the pipe.

I would use something flexible rather than epoxy. Pressure and flexing will likely peel the cable jacket away from rigid epoxy, whereas something flexible like a silicone or a urethane will follow the flexing of the cable. I have another colleague working on the Manta 5

and they are having exactly this problem with an epoxy based cable seal.

Which ever cable you go with, see if you can find a jelly (gel) filled version. This will be more resistant to pressure, and also stops the issue of water creeping up the cable. These types of cables are rated for permanent burial, where water ingress is an issue.

1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob