Dear all,

had a few thoughts on how to build a simple ROV out of more or less standard material.

Before I hit the order buttons on ebay I was wondering if you could give me some feedback on my ideas.

Target is to have a ROV that can operate down to 20m depth. I want it to be RC controlled using a float on the water surface having the receiver and a 5.8 GHz transmitter for video (VTX). Onboard there should be a small camera with "livestream" to the VTX and finally to my video goggles. I am planning to maneuver using 3 brushless motors (2 in the rear, 1 on top). Battery will be on board. Pressure sensor, sonar and OSD can be added after a few successful trials an reaching the targeted depth. I believe electronics can be done, having some experience coming from quadcopters, adruino, ...

What concerns me is the frame / body of the ROV. I am wondering if a plexiglas body will do it. Targeting to reduce volume as much as possible, I had the idea to use a 80mm/70mm (outer/inner diameter) tube with 10mm thick plexiglas disks glued to both ends. The main tube will be made out of two parts, the front part with 250mm and the rear part with 100mm length. I will put a "connector" tube inbetween, with 70mm/60mm and a length of 100mm. The picture below is showing the idea. My question is: will this withstand the pressure in 20m? Can a wide plastic tape around the intersection between the two parts be used to seal the body?

I appreciate your feedback!

Ole

https://drive.google.com/open?id=18z...5Tm56uK5Jb1dIQ

I don't think your idea of a piece of tape will be enough to seal out the water. Remember, pressure increases 14.7psi every 32 ft of depth, so you'll have approx 30psi of water pressure trying to get into your WTC at 20m depth. That's why the majority of hobby ROV's available use o-rings to seal the WTC. The water pressure forces/crushes the silicone o-ring against the square-cut groove and that seals it up tight.

It would be the same as trying to use a piece of tape to seal a hole in a fully-inflated bicycle tube. That would be 30psi trying to get out. In my experience, it just leaks out.

I can see using a float on the surface, but how are you going to control the ROV? Will you be able to send PPM signals from the RC receiver to the servos/esc's down 20m of cat5? Most ROV's use a rs485 serial connection from the surface.

I had thought of using a float for my ROV (a Big float), but if it got caught i'd need a boat for retreival anyway. Have you considered using a kayak topside with you in it?

BTW, until jan 31st Banggood has a sale on tiny cameras. Eachine 1000 tv lines, 1/3" ccd imager, wdr, 110* fov for only $9.59!
https://www.banggood.com/Eachine-100...r_warehouse=CN

BlueRobotics sells a 4" WTC 12" long for $305. It has a hemispherical dome on one end and a cable-penetrator aluminum plate opposite.

I have seen the blue robotics page. Actually, l want to have the rov for <300 USD so I am wondering if I should try some new approaches.

Please note, as pressure inside the hull is lower, the tape will only be pressed harder to the tube. This is different to the example with the bicycle tube.

It would be like putting tape on the inside of a bicycle tube... I don't think that's a good idea.

As for the float - any concerns with sending the RX / vTX signal 20m without losing the signal? I want to do something similar, but I am afraid that the drop will be too much for the tether without some sort of booster... Let me know what you think!

My goal would to use a FC from an old quadcopter to use OSD to display orientation. I don't know what would be better: VTx on the ROV or the raft.

This is the cheapest 3-dof ROV i can think of. Build the frame from 3/4" pvc. Paint the pvc yellow for visibility and to protect it from the UV rays of the sun. Use 4+ empty freon tanks for bouyancy (an HVAC guy will give them to you). Use red building bricks for ballast.

El-Cheapo
___________________________________
Chartsea 4300KV Brushless Motor/60A ESC/Program Card Combo $40 x4 ($160) Amazon

Freewing 90mm 6 Blade Fan $8 x4 ($32) MotionRC

2 Rmilec ppm/pwm/sbus signal converters $14 ($28)

10k pot 2-axis joystick $12

(5) 10k pot micro joysticks $6

18 Servo Testers @ $2.64 ea.
https://www.ebay.com/itm/Professiona....c100005.m1851

10pcs 12V T10 Wedge 24SMD LED Light Bulbs $12 https://www.ebay.com/itm/10pcs-ultra....c100005.m1851

2-pack 12vdc 10Ah SLA batteries (x2) $80 Amazon

12 video baluns $13 eBay

500ft Cat6 outdoor 100% Copper $90 Amazon

Eachine 1000tvl microcam $13 Banggood

$493.52
___________________________________
Pull the pots off the servo testers. I included the link for those because you'll need the ones you can physically switch to 'manual', the newer ones auto-select 'automatic' when power is applied.
Use the 10k pot joysticks you bought, solder the pots to the servo testers. This will give you 18 'channels'.
Plug the 18 servo testers into the Rmilec signal converter. It will convert the PPM output of the servo testers into an Sbus signal which will go through the Cat6 via just two wires to the second signal converter, set to mode 3.
That will convert the Sbus signal to individual PPM outputs. (Like using RC Tx/Rx sets via DSC (direct servo control).
Use the PPM outputs to drive LED's, ESC's for thrust, servos, etc.
Chop 3 blades off the 6-blade EDF prop for best efficiency in water. The 90mm prop is 3.7" in diameter, so use a 3" long piece of 4" PVC for the duct. Put wire mesh over the ducts to keep weeds and fingers out of the props.
You could try coating the ESC's with Epoxy resin fiberglass to waterproof them. Try one first, if you fry it you'll have to put them in watertight containers.
Coating the LED's in epoxy will waterproof them just fine.
Using video baluns between your camera and topside monitor will allow you to run the signal through 2 wires over long distances without any loss of video quality.

Something similar to this. Although i would move the horizontal thrusters forward from the rear, to the middle, either side. This will give better yaw response. Remember, the bigger you build it, the more stable your video will be.

Homemade ROV (2 min 37 sec)

Self amalgamating tape (the stuff sold in auto part shops as radiator pipe temporary repair tape), properly applied, will seal pipe to pipe joins provided that they don't involve any angles. Worked just fine for underground cables in areas liable to flooding because the pressure was sealing the tape into the join.
Confused by all the servo testers unless there is a need to control manually from the surface float.
No experience with using S-bus, but thinking about it, -
If the receiver doesn't do S-Bus, it needs an adaptor to convert to allow a single cable pair.
At the ROV end, a matching demultiplexer,
To go 20 metres might need some sort of booster, possibly at both ends. At the send end to ensure that a reasonably noise free signal will arrive, at the receive end to ensure that the receiving device has enough signal. Probably a pair of 741 op-amps, or whatever has taken their place these days.
Sending the TV signal back up is a different problem, and will probably need some sort of booster to get a viable signal back to the surface via what will probably be a lossy cable.

Originally Posted by mfr02

Sending the TV signal back up is a different problem, and will probably need some sort of booster to get a viable signal back to the surface via what will probably be a lossy cable.

Nope, that's what the video baluns are for. Perfectly clean video over twisted pair cable.

And i guess if he has a receiver with sbus-out on the float/boat, he won't need all the servo testers. I just included them to provide a ppm-output from joysticks.

I just can't fathom (pun) building a decent ROV for under 300. So i was trying to help the guy get one built for close to that.

Come to think of it, i saw some FlexSeal tape like you describe at WalMart recently.

I wasn't thinking quite on the ball with the TV signal - obviously its at whatever the camera is kicking out, not at transmitted frequency.
Still, 20 metres is quite a way, a boost at source might drop the signal to noise ratio to acceptable levels. It was an outside broadcast trick to add quite a lot of decibels to the send end of a link where noise might be a problem, then attenuate it back down (along with any noise) at the other end of the link. But that was watching the BBC radio OB lads when I was "Lineman liaision" when they rolled into town, so they were sending sound down a line, hoping for studio quality at the other end.