Simple way to increase the depth capability of Atmospheric Housings


Many people forget the biggest problem with atmospheric housings is not the pressure, it's the "differential" pressure. Any housing will have a psi rating for which it will "collapse" or "implode." One simple and inexpensive way to increase this depth rating is to pressure the atmospheric housing up with a compressed gas while the ROV is on the deck.

For example, the psi of water pressure pushing on the atm housing at 100m of water depth is about 160.5 psi. Assume the crush pressure of the housing to be around 200 psi. Now if you were to pressurize the housing with air or even better dry nitrogen to about 90 psi before you started the dive, you would effectively add an additional 50m of depth capability to your ROV. (50m water depth is about 87.5 psi) When the ROV reaches about 50m of depth, the internal gas pressure would equal the external ambient pressure.

There are concerns in using this method though as you have to ensure you do not have any "hollow" electrical components that might implode at the ~90 psi over pressure. Also, you have to ensure your "seals" on the atmospheric housing work in an outward force as well as an inward, which most do not.

I just found this site today, thanks to kick starter. I hope to build my own OpenROV (actually 3, one for myself and one for each of my children who are 11 and 10.) I will be around frequently until I depart for the South Pacific for work in a month or two. I plan to work on a power system for this project that will allow you to trickle low current power to the ROV continuously, and use the power as needed. :-D Worked on a similar project for one of the companies I worked for, but the project got canned because they didn't need the technology anymore. Granted, that was to run a 200hp ROV so it was a bit more complicated. LOL

Glad to be here and look forward to getting to know you all.



This is a very good idea. How about fitting a standard bike/car tyre valve to the capsule? Bike pumps and pressure gauges are very cheap.

If the ROV is never operated deeper than the limit to which it has been pressurised it need only be designed for positive pressure inside, not both positive and negative.

The BIG advantage of this method is that if it leaks then it's air leaking out, not water leaking in and the electronics don't get wet. With a bit of luck you'll see the bubbles as soon as you launch it. Up to a point, the deeper you go the less likely it is to leak!

On the other hand, an exploding capsule on the surface is a lot more dangerous than an imploding one under water, so it had better be strong enough!



Hey man,

I think the threads regarding filling the tank with mineral oil are discussing this.


Just use a spare dive regulator. It puts out ambient pressure by design and even has a vent valve for assent. The first stage reg already fits the bottle. Air consumption would be relative to how often you change depth up and back down.

Since you now have a supply of compressed gas on board, you have the opportunity to do active buoyancy compensation. This comes in handy if you intend to collect anything while you are down there.



This is a good idea but can really be improved upon to go to unlimited depths theoretically. All one needs is an electrolyzer to convert water to oxygen and hydrogen. This would provide the necessary pressure inside the housing to resist outside pressure. Take a look at this electrolyzer available from

AT $180 for a 5-pack we are talking about $36 per piece. Of course, you need a source of pure water and a cheap reverse osmosis membrane can do that.

You could go to any depth you wanted since nitrogen, oxygen and hydrogen will never liquefy. The critical temperature for these gases is way below 0.

When you are coming up, you would need to let off gases, like a pressure cooker does. Perhaps a simple valve like one of these will do.

Pressure Relief Valve with G 3/4" pipe thread. ABS Pressure Relief valve with G3/4" thread. A low profile relief valve opening when a predetermined pressure is reached. Nominal opening pressure 0.07 Bar.

This valve (at less than $10 a pop) could also be used as the water intake valve for the electrolyzer.

So if power was provided by a long neutral bouyancy tether and the ROV did not descend too fast, it should be able to go to any depth necessary. I guess the only caveat is that there should be no spark inside the housing.


PS. You probably need a differential pressure sensor to control this system.


Wouldn’t it be simplest to put a small CO2 cartridge inside the housing?

And have a mechanical differential pressure sensor use this to equalize the pressure.


You could do that. But below 31C and, I believe, 2,500 feet of water, the CO2 would liquefy and the ROV may not survive. Even then, a great improvement in depth.



whilst theoretically all correct, pls BEWARE OF HIGH PRESSURE GAS FILLED EQUIPMENT. You do not want to be around it when the ROV comes back to the surface and the overpressure valves did not release the pressure correctly. It might disintegrate very very fast and unlike an underwater implosion, this time it will be an explosion. :-) Gas expansion has a little trickier behaviour as liquids... :-)

Better solution is to do what James Cameron did for his National Geographic live feed from the bottom of the titanic. (all available on youtube). Whilst he was sitting in the MIR1(2) deepsea underwater vehicle outside of the titanic, he used two remote ROV's (designed by his brother) that where only slightly bigger than the openrov's to fly right inside. His ROV's where oil filled. Only problem here, you need a lot of x ray to identify all electronic components to see if they have any air bubbles and are cabable of the pressure. (I guess that is the reason they where so expensive. I hears somewhere that they had cost a lot of $$) He used fibre optic cable that reeled out from the rov and had an autoattach function back to the base on the MIR. Meaning he could fly through the wreck and didnt have to come back out where he went in... just go back to base, have the manipulator cut the small fibre optic cable and reconnect.... and ready for next flythrough...(still aways away, but Im working on that technique....).

As for "normal" depth rating I would always go for standard housing. There is actually a quite amusing video on youtube where a guy has posted his DIY ROV project with a depth rating of 2.000m. (search for sv seeker and rov)