Arduino under high pressure - oil filled pressure compensated housing - ROV

Hi,
I am putting my ROV together and thinking about having a oil (none conductive) filled pressure compensated housing for the electronics.

Are there any components on the Arduino Mega 2560 Rev3 that would collapse under high pressure? Can I replace them with other components to make it work?

Thank you for reading

There are a couple of electrolytic capacitors near the barrel plug that may have issues with high pressure. If you're using a stable power source they may be redundant.

I'm not too sure about the crystal oscillator either. This is only needed for the USB interface though. It could be replaced with a socket so you only plug the oscillator into the circuit when you use the USB (when not submerged)

Oh and it's just possible that the reset button will be triggered by the pressure. Obviously you could simply remove this and replace its function by shorting the reset pin to GND if you want to reset the board.

Thanks for a very informative answer.

Replacing the oscillator with a socket and only connecting it when I need USB interface seems like a simple well working fix for that component.

Removing the reset button is also a quick and easy fix.

Now to the electrolytic capacitors, you say that they may be redundant if I use a stable power source. Are there any other components that I can replace these with? I’m only just reading up on Wikipedia but could a capacitor with solid electrolytes be an option?

I would think it would take many atmospheres to crush the crystal casing considering its size - possibly in the order of 1000.

The reset switch will not be actuated by pressure - it is open - the oil will simply seep directly into the mechanism. It may or may not work too well afterward.

Not sure whether the electrolytics have significant voids or not. Other components may also have voids - who knows?

Paul__B: I would think it would take many atmospheres to crush the crystal casing considering its size - possibly in the order of 1000.

Possibly, would you be sure that it would be OK to a few hundred?

The reset switch will not be actuated by pressure - it is open

Not necassarily. It's open at the top, the base is a solid piece of plastic with just the legs projecting through. There's a cylindrical well inside with a cup shaped piece of spring steel. (nickel plated). On top of that is a plastic bush that fits quite nicely within that well. It's not designed to be air tight, but could (especially with the addition of a viscous fluid) produce a close enough fit to form a compressible air bubble in there.

Not sure whether the electrolytics have significant voids or not. Other components may also have voids - who knows?

It's not just voids that are a problem. The electrolytes are generally compressible. This could allow sufficient distortion of the can to damage the dielectric. I'd feel more confident with tantalum beads, but even then, with the pressures anticipated, I think a bit of research would be worthwhile.

I doubt anything else on the board would have a problem.

What about heat dissipation ?ing viscocity

Convection circuits do not work well with increasing viscocity

Boardburner2: What about heat dissipation ?ing viscocity

Convection circuits do not work well with increasing viscocity

It's sitting in the ocean and you're worried about heat dissipation :fearful:

Just playing devils advocate, why not just stick the processor environment inside a little pressure proof housing and stick a waterproof connector (fischer etc) to the rest? The housing would be small enough that the forces would be manageable and you'd have the advantage that any software reloading etc wouldnt involve a pile of clean up first :)

Out of interest, what sort of pressure rating are you looking for?

I'm aiming for 300m depth to start with, pressure proof housing at this depth isn't that big of a deal. I am just trying to be one step ahead when I finally want to push it further. I am also aiming for a light portable vessel and using pressure compensated housings will allow me to do that.

It's also a matter of hull penetrations, they are quite pricey so I would prefer making my own. Making those to withstand really high pressure seems to be quite difficult (now I am thinking ahead, deep, deep).

I haven't really decided yet, the big drawback is that it is messy when I need to change a component or something like that. But I'm starting to believe there are more pros then cons.

One approach could be to make a pressure proof setup to start of with that can reach around 300m and then convert that to a pressure compensated setup when I want to go deeper.

Yep, the quest for a reasonably priced connector is eternal ;) If you do find any, feel free to link them :)

I guess if you're using a non-conductive oil, you dont worry too much about clean up, just plug and go.

a quick solution might be to get something like a VR3 (dive computer) second hand on fleabay and strip out the internals so you can reuse the housing/seals/connectors.

KenF: It's sitting in the ocean and you're worried about heat dissipation :fearful:

It may be an issue ken, if the oil cannot transfer the heat out, if it acts like an insulator. I know some of the oil based cooling immersion tanks for PC's (Ie stick a PC sans case in a fish tank and fill it with oil) required pumps to circulate to oil to facilitate cooling.

OP, Do you have a pressure vessel to test your pressure housings and penetrations? I'd be interested in seeing your setups, please keep us updated.

I haven’t built anything yet to pressure test housings and penetrations, I haven’t decided yet if I should do that or build by theory and calculations. I will probably do live tests in the field.

Jaman42: I haven't built anything yet to pressure test housings and penetrations, I haven't decided yet if I should do that or build by theory and calculations. I will probably do live tests in the field.

It's actually pretty difficult to reproduce the pressures you'd be expecting without actually dropping something in the sea.

It might be interesting to just drop an UNO overboard on a line and let it sink a few hundred feet, then pull it up, wash and dry it and just see what components have failed.

If you already have the oil, expanding upon Ken's idea, put some and an UNO in a baggy (designed to stay closed under the circumstances) and drop that over the side, that way you can (hopefully) eliminate water damage as a source of failure.

KenF: It's sitting in the ocean and you're worried about heat dissipation :fearful:

An aquintance who builds model subs has this issue. I fly and dont

Oil is actually a good insulator unless its moving. Thats why you see the big loops on electricity substations.

Large mass gets a thermosyphon going

Marmotjr: If you already have the oil, expanding upon Ken's idea, put some and an UNO in a baggy (designed to stay closed under the circumstances) and drop that over the side, that way you can (hopefully) eliminate water damage as a source of failure.

Yes i think that would be the simplest, i would powerit up though to test the possible temperature problem.

What oil though , transformer oul ?

Mineral oil seems to be widely used for this purpose, not sure whats best thou.

Connectors, why bother.

Oil floats.

Leave enclosure open at bottom and just poke wire in

Doh.

Just realised the cable itself would have to be filled with something the same density as water for that to work.

ROV cables must be something special i guess.

KenF: It's not just voids that are a problem. The electrolytes are generally compressible.

Is this true with solid electrolytes as well?