ROV design

Translate This Thread From English to

Threaded View
Hello,

I am working on a ROV that will hopefully clean ducts. I'm going to
have 2 variable speed drive motors, a turret motor that I can control
the position of, a variable speed motor for a rotating brush, and
maybe some sort of cleaning agent sprayer. It will also have a camera
that sends video information back to the operator.

My problem is that I want the cable between the ROV and operator panel
to consist of just two power wires and a communication cable. Does
anyone have any suggestions on a low cost method to send the user
outputs and video back and forth on a single communication cable
(maybe a CAT5 cable)?

Thanks, any ideas are appreciated,

Joe McKibben


Re: ROV design


make the cat5 a crossover

put a mini ITX motherboard on your robot

have a laptop computer at the other end of the cable.

run a video server on the bot.

Rich


Re: ROV design


That is an interesting idea. But a little more expensive than I was
thinking of.

Thanks,
Joe McKibben


Re: ROV design




Joe McKibben wrote:

It's far cheaper than the alternatives you are talking about.

Used pentium II laptop: $50 on eBay

Mini-ITX board: $100

Webcam: $40

Your 4 variable speed drive motors and controllers, tracks,
brush, sprayer, etc. will cost *far* more.

In addition, it won't actually clean ducts.  You will be
spraying the dirt and then moving it around with the brush.

Free Clue: if your robot comes out of the duct and you don't
have to empty out a couple of pounds of dirt, the dirt is still
in the ducts.

 






Re: ROV design


Thanks for catching me on this. But I just didn't mention it would be
connected to a commercially available duct cleaning vacuum system that
has a spinning brush on it. The robot would carry this up the duct and
also have the chemical sprayer.

Joe McKibben



Re: ROV design


I do understand how the laptop would save me money by not having to
use the case and all of the mechanical pots., switches, also it would
have a screen already.
It would be pretty compact also.

But I am wanting to make more than just one of these robots. So,
hoping on good laptop prices on E-bay isn't going to help me with
this.
Also, I want the system to be more robust and be able to withstand a
little abuse. I don't know how many drops a laptop might take.

The communication between the robot and operator panel I was hoping to
base on a couple inexpensive ICs.

Thanks for your ideas. It gives me something to chew on.

Joe McKibben


Re: ROV design



Video: Camera baluns on each end the type thats used for CCTV

Outputs depend on what they are...Protocol, voltage etc etc

I wouldn't send power and (outputs and video) over the same cable as you'll
get interference. Look at using two cables with one shielded. As far as I
know there isn't a cable where separate conductors are shielded though I
must admit I haven't looked.....



Re: ROV design


What I am probably going to do is get a custom cable made up, it will
probably have two or three cables in it. There are some companies that
do it.
The video camera will be like a back up camera for cars.

Joe McKibben


Re: ROV design


The way I am going to control it is with a two-axis potentiometer
joystick for the drive, probably a dial for turret position and a dial
for brush speed. A single switch for the spray. It will all be housed
in one of those plastic briefcase things. There will be an LCD for
viewing the video from the robot.

So, I need a way to gather all of my analog(4) and digital(1) signals
in to a single serial signal that I can convert into the PWM and Relay
signals I need on hte robot.

Joe McKibben


Re: ROV design



Stick a uC at either end and use RS-232.

Remember voltage drops over wire considerably if you've got a long
length.........



Re: ROV design


Have you considered RF? and eliminating the cable?

XBee might do well, although you'd probably have to put an antenna
inside the duct at the point of entry. I'm not certain, but I'm
thinking the inside of the duct might act as a fair conduit, or wave
chamber, getting the signal well distributed everywhere inside the
metal ducting. A XBee link would take care of bidirectional command
and reporting. However, your video is another matter. At XBee speeds,
you'd have a picture every ~2 or 3 seconds, and it would dominate the
whole channel/bandwidth to get that. 801.11 might make a good path for
video, and maybe control as well.

At those high frequencies it could go either way, a really tight
enclosed wave guide, or a squelching faraday cage around ever corner.
A little experimentation might tell you which.

Randy


Re: ROV design


When I first was given the idea for this project that was my first
inclination. But the person who gave me this project thinks it will
interfere   with wirless communications in hospitals were he does a
lot of duct cleaning. Also though the robot is going to be carrying a
vacuum hose so carrying a cable along with it probably will not hurt
to much.

Joe McKibben


Re: ROV design




RMDumse wrote:


That's a heck of an RF transmitter if it can eliminate a caple that
supplies power to run several electric motors...


Re: ROV design


That would be wouldn't it. That much RF, and you could just burn out
the dust off the metal surfaces, and skip the whole ROV thing, huh?

No, I meant the communications cable. The power cable or alternative
of batteries would still be required.

Randy


Re: ROV design


Yeah if you used RF you would have to use a battery or still a cable
but just with the power wires.

Joe McKibben


Re: ROV design

On Sat, 13 Oct 2007 14:58:19 -0000, Joe McKibben


Duct/pipe bots seem to be popular things to try to construct.
From my simple experience, you probably could control it from a
laptop thru a single piece of cat3 phone wire. One conductor for
rs232 to a controller, one conductor for video, one conductor for
+ power for the controller, cam, and servos, and a common ground
(the bot would need to have batterys for the large motors, lights
and such). Would be easy to test. From what I've seen of the
homebrew duct/pipe bots, I think most were eventually abandoned
for various reasons.

Re: ROV design


I like this idea of carrying a battery on board so I wouldn't have to
use larger gauge power wires in the cable. But, I already have some
pretty tight space requirements for the robot.

Thanks,
Joe McKibben



Re: ROV design



I personally would use an on-board battery, and run everything else through
a single CAT-5.

Oh, and use RS-485, not RS-232. You need an RS-485 transceiver at each end,
but you can run RS-485 reliably through thousands of feet of cable.

Later,
Jon

--------------------------------------------------------------
   Jon Hylands      Jon@huv.com      http://www.huv.com/jon

  Project: Micro Raptor (Small Biped Velociraptor Robot)
           http://www.huv.com/blog

Re: ROV design


I did some research on RS-485 and found some Maxim RS-485 tranceiver
ICs. I get how they comunicate between each other but how do I add my
Outputs to the circuit?
Also could I transmit the video over the RS-485?

Thanks for the help,
Joe McKibben


Re: ROV design


I'm not sure RS-485 has enough bandwidth for video.

100MB ethernet has the required bandwidth for video.

-Wayne

Site Timeline