How much pressure can the electronics in a computer withstand?

[Pleiades]

I’m thinking of building my own ROV (remotely operated vehicle) which is basically a remote control submarine.

 

Ideally I’d like the ROV to be able to withstand going fairly deep, say 200 feet, which means it will need to withstand high pressure (6 atmospheres / 88psi / 608kPa). I was thinking rather than build the ROV out of materials that can withstand such pressures, why not fill it with an uncompressible yet inert liquid, like oil?

 

Well, one reason not to do this is that the electronics inside the ROV might break under the pressure anyway. At the very least I’ll have a video camera in there (lets not worry about the oil changing the optical properties of the camera just yet) along with some motors for propulsion and I may also want to have a small computer in there. So my question is: can anyone think of a component in a computer or other complex electronic device that would not be able to withstand such pressures? I’m a little worried about the capacitors.

 

Thanks in advance.

[YT2095]

interesting!

 

you`re correct about the capacitors, the electrolytic types anyway, they could quite easily be crushed, however tantalum capacitors will be able to take the pressure much better.

the problem comes from Air pockets, even in the resin surrounding an IC (that`s whats responsible for bits flying off when over heated, air expansion).

the other problem will be operating temperatures more so than the pressures involved. high stab mil spec parts should work ok though.

your Oil idea is a good one! but I wouldn`t use it around RF stages and components as it will adversely affect capacitance and possibly inductance, so any radio circuitry will have to be treated differently, perhaps a silicone then a resin outer shell?

 

let me know how you get on with this, you`ve got my Interest here

[Pleiades]

Glad I’ve piqued your curiosity. I may end up not using a computer at all but let’s entertain the idea for a while.

 

The capacitors can be replaced with matching ones of the type you mention, so that’s no biggie. The problem with the ICs did cross my mind, but now you’ve got me worried, how big an air pocket are we talking? Will it allow the resin to deform enough to crush the core? Can the resin be replaced with something more flexible without damaging it?

 

I don’t think there will be a problem with overheating; I’d be using a low power computer (something like this: http://www.mini-itx.com/store/?c=2#p1601) so the heat it generates should be able to conduct through the oil, which conducts heat better than air, but worse than water. If it’s a problem, I can make some sort of metal “heat pipe” that conducts the heat into the ambient water.

 

What do you mean by radio circuitry? If you mean something like Wi-fi or Bluetooth, I’m not planning on using anything like that (wouldn’t get too far through the water). I don’t know much about choke coils, but would they be affected by the oil?

[YT2095]

for the ICs there`s nothing you can do really, if it`s in the black resin/plastic packaging already, there`s nothing anyone can do to alter it, I`m not sure how compression forces will affect them at all, I know expansion can be quite devastating though.

if you`re not using any form of radio link (I`de go for low freq anyway), then capacitance and inductance won`t be an issue

Overheating isn`t a prob obviously, but there is a Low temp threshold with certain parts, again, you oil may equalise this and it shouldn`t be a real issue unless you plan on going Arctic with it

the worst offenders in the IC range are things like older CPUs where there`s a metal plate over the chip itself (often gold color), the IC body is a ceramic and will take the pressure no problem, but the metal plate/cover COULD deform and bend inwards, again, there`s nothing you can do about this, also EPROMS have a void just bellow the erase window, so be carefull with them too.

[Pleiades]

Another cool, but probably unfeasible idea id like to try is some sort of deep sea probe that would be lowered about 2 miles (3.2km) down, and would have some kind of video recording device, and some lights. Somehow I don’t think I’m going to find off the shelf parts that can withstand 4700psi (32424kPa) at 32F (0C), but it’s a cool idea nonetheless. Apparently alkaline batteries can be modified to withstand 10,000psi, and freezing temperatures, so that’s ok (http://pergatory.mit.edu/rcortesi/portf/highpresbat/highpresbat.htm).

 

Well, there is only one way to find out about the ICs and that’s to do some testing, and it could be several months before I get the opportunity to get out into deep water.

 

The low temperature will be an issue for the deep sea probe, but not for the ROV since the water temp at 200 feet is nothing extreme in my area.

 

If I do use a computer in the ROV, it’s going to need solid state storage, probably a Compact Flash card; I assume EEPROMs will be ok because they don’t have an erase window.

 

While I’d love to make the ROV completely wireless I don’t think its going to work. There are 2 ways that I know of to have wireless communication underwater: acoustic ultrasound modems and ELF/VLF radio. Both of these technologies have extremely small data capacities which are no where near enough for video, one acoustic modem I looked at has great range, but only transmits at 15 kilobits/second, about 1/3rd as slow as a 56k computer modem. VLF radio is even slower. A blind ROV is worthless and I can’t think of a way to do wireless video underwater. This is not unexpected, as all the ROVs I’ve looked at use a tether to link them to the surface.

 

Time to do some testing I guess.

[Externet]

Hi.

Use paraffin as a filler in a box sorrounding the electronics instead.

And consider a tethered robot if you want real time video.

Miguel

[padren]

Well, you'd have to be very very sure you had all your components juuust right but you could also try to pour some sort of clear resin that hardens solid to box in your electronics, then pour oil around that block to deal with the remaining air.

 

My thinking is, the pressure would be distributed throughout the very hard resin, allowing very small voids of where your air pockets in the electronics to exist at 1 atmos of pressure.

 

Oil will transfer that pressure directly against any surface that has a void trapped in it, but a very hard resin - I think - would help nullify that.

 

Its just an idea though and may be flawed, do you think it may work?

[Pleiades]

I think the resin idea could help. The air pockets found in ICs will probably be square, and a hollow box will be easily crushed by pressure, but a hollow sphere can withstand quite a bit more. Forming a more spherical enclosure around the air pocket would give it a more uniform strength.

 

Of course, it may turn out that some ICs have no hollow pockets at all. As soon as I decide on a design and control system for this ROV, I can test out how the electronics will withstand the pressure.

 

My biggest stumbling block right now is deciding what kind of propulsion/navigational design to use. I thought about using a rudder and hydroplane system like on a large submarine, but this doesn’t allow the ROV to maneuver without going forward or backward. The more conventional ROV design is to use a system of thrusters. After some thinking I decided the best thruster system is to use a reversible thruster in a tube that goes through one side of the ROV and out the other (ducted thrusters), you would need 4 such thrusters and 1 main drive propeller. This idea is nothing new, and after some research I found that the navy has something similar to what I was thinking of. The DSRV propulsion system appears to be a good mix of maneuverability and top speed. http://www.fas.org/man/dod-101/sys/ship/dsrv.htm

 

The DSRV has a shroud or nozzle around its propeller; it can tilt on 2 axes, allowing it to steer the vehicle up, down, left and right. While this is a nice feature, I think the thrusters can achieve the same effect without adding complexity to the project.

[lightwave]

An ROV is a tethered vehicle. You appear to want to build an AUV which is an Autonomous Underwater Vehicle - meaning that it is untethered and under it's own control. It's quite usual for AUV's to be always moving. After all if you want a vehicle to be stopped then just tether it. AUV's therefore always have propellers like conventional submarines.

 

All AUV's have an air space surrounding the electronics. You cannot reliably surround electronics with oil, however electrically inert it is. In any case you need some bouyancy available and for that you will need an airspace. No AUV has no computers. They all have several computers. Usually one for navigation, one for control and one for the sensor package (instruments). You may also need one for communications.

 

You can get video back from an AUV without a cable. Control is available to you through acoustic modems. The data rate only needs to be 300 bps for control functions, after all you aren't performing maneuvers at 100 knots. Typically an AUV will move at a speed of around 4 to 8 knots. It's not fast, but if you want faster you'll need to have many more batteries and a bigger hull.

 

Mostly they use lithium-ion batteries, but there are still a few who use lead-acid cells. Both will tolerate pressures to 300 metres. Most of the research and industrial uses all occur in the top 100 metres of water. To dive deeper has no purpose really unless you want to follow seals around the place or examine deep ocean objects. The deeper you want to go the deeper your pockets will need to be. The pressure imposes huge costs on the hull building. A 1000 metre dive will mean a very sophisticated vessel which might cost $200,000 to build.

 

Your biggest problem in building an AUV is the hull. The electronics is straight forward to do. Writing the software is a tough one. Navigation is a specially difficult problem as you'll need a 2 axis or 3 axis magnetometer and writing the software for that is not easy. The hydraulics (shifting water around) on board are not difficult but they do need to be 100% reliable. The electricals are relatively easy.

 

For the best AUV's they have a notebook computer on the surface communicating with the vessel. You'll need software to run on the surface computer to enable monitoring and control. It's usual to want to know battery volts, speed of vessel, depth, location, heading, height, motor revs, interior pressure and sensor values. You can do all this at only 300 bps, it's not really that much data.

 

Hydroplanes are always used to control depth. All you need for those is modest sized stepper motors. Most AUV's will have a maximum mission time of 2 hours but there's a few that will run for many more hours. Survivability is a problem. If your vessel sinks you'll need to be able to recover it from possibly a big depth. If it runs off on you it'll be lost for ever as it could be anywhere inside a 10 km radius circle - about 40 square km. It'd cost hundreds of thousands to recover it.

 

The vessel needs 'fail-safe' systems. If something goes wrong it auto-surfaces and signals it's position, often with a GPS and an optical beacon. Recovery at night, before the tides get it, is a problem.

 

Hope that answers some basic questions. If you have any more feel free to PM me.

[calbiterol]

I was thinking rather than build the ROV out of materials that can withstand such pressures, why not fill it with an uncompressible yet inert liquid, like oil?

 

I understand the desire to think "outside the bowl," but wouldn't the former be much easier? A simple pressure-proof, openable case can be easily made from PVC piping. That's actually an interesting idea - create a miniature submarine out of PVC. I might add that to my list of projects.

 

Anyway, I didn't have time to read past the first two posts or so, so I have no idea if this has been explained or whatnot already. At any rate, thought I'd suggest it.

 

Cheers!

[Pleiades]

You cannot reliably surround electronics with oil, however electrically inert it is.

Umm… Why not?

 

 

You can get video back from an AUV without a cable.

If you know of some way to transmit live video through the water in real-time, please, let me know.

 

This thing is a toy, I want to be able to control it remotely, I don’t want it to do things on its own, I want to control it; I want an ROV, not an AUV.

[lightwave]

If all you want is a little toy sub there's loads of them on the net from $25.

 

I designed an AUV some time ago and I produced a means of getting real-time video back to the surface without cables. I will not disclose how I did this.

 

If you just want to drop a box in the water which is tethered and sends video back to the surface that is easy to do. There's designs on the net to do this and all the components are readily available.

 

When something is tethered, you don't need to pay much attention to the hydrodynamics of the vessel. It can be any shape. To make it go somewhere while hanging off a cable you'll need some thrusters. All you need to do is have an electronic means of turning them off and on as required. It's usual to use power FET transistors for that. Your tether cable will have video cable, control lines, stainless steel core and power cable.

 

Although it's easy to design a controller, there's off-the-shelf items to do this which are very cheap and come from an unusual source. They can handle 25 amps easily and cost around $35.

 

If it's tethered, you won't need any system of diving the vessel, just allow it to sink. You might have to apply lots of ballast to get it to sink.

 

One of your problems is that the cable has a high mass and a tiny vessel can't tow it. So your field of radius might be small if your vessel is under- powered. The deeper the dive the more powerful the motor will need to be to get a reasonable field of radius.

 

The cable might have a mass of 500 g per metre. A 100 metre cable may have a mass of 50 kg. You will have to have a poweful motor/motors to tow it around. Just guessing, but around 110 watts seems practical. 110 watts at 12 volts is 9 amps. You'll need a small car battery. For a 100 metre dive expect the mass of the vehicle to be around 100 kg.

 

But if all you want to be is a 'bathtub warrior' then just spend your $25.

[insane_alien]

i've surrounded electronics in oil without any adverse effects before. in fact it was quite beneficial as i no longer needed a cooling system for the cpu (pentium III 500MHz) and motherboard that was submerged in the oil.

[RICHARDBATTY]

Would it be possible to use multiple channel sonic comunication, then reassemble the data ?.

[lightwave]

I think it would be impossible to use multiple channels. The conductor (the ocean) has multiple pathways and multiple delays so each packet of data would arrive at very odd times. There would be multiple requests to re-send data (as there is anyway) on a multiple number of channels. Acoustic spread spectrum has been examined from time to time, but never used (commercially) that I know of.

 

Data rates are slow through water, but they get even worse when the vessel gets a long distance from the reciever.

 

Whales can manage to be heard over distances of 2500 km. But they don't transmit packets of data - they sing their message using tones. Inherently their 'data rate' is very slow and despite the distances they can still understand each other but there's a massive 'computer' at each end of the whale's transmission path - their brains.

 

Morse is a very reliable communications system, it's very slow, but is still today regarded as the most reliable system of electronic communication we've ever invented. It's reliable because the operators have a powerful 'computer' at each end.

 

Speed requirements for data transmission underwater is very high. 300 bps is adequate for basic data about functions on the craft, but video requires MUCH higher speeds. Except I have a method where the data speed requirement is much lower.

 

It's often said that the ocean is very noisy now (due to ships) and whales can't hear each other. In fact whales can still hear each other over vast distances due to the range of tonal sounds they make.

 

I believe whales can hear infrasonically. They would hear every tidal wave and undersea earthquake probably as a great rumbling noise (?).

 

Rain can be heard on the surface of the water for many kilometres.

 

Vessels can be heard and even exactly identified for hundreds of km's. Every sub has a signature noise, which navies like to keep track of.

[lightwave]

I've just been doing a bit of research and it IS possible to use spread spectrum on acoustic transmissions. There are a few commercial products out there. Although they are not small and they are not cheap.

 

A tethered ROV doesn't need an acoustic modem, all the comms are via it's cable of course.

 

But modellers use a 27 MHz or 40 MHz Radio Control transmitter to connect to their subs. In fresh water the transmission, only about 500 mW, reaches about 2 metres under water for short distances.

[YT2095]

I think what Richard was on about was a single data stream but instead of carrying On/Off data (binary), using something closer to DTMF tones, that way you could send more complex data over a single channel.

[Pleiades]

I found this article which sheds some light: http://www17.tomshardware.com/2006/01/09/strip_out_the_fans/. These guys found that the oil caused the computer to crash if it got into the CPU socket. There explanation is that because the oil works as a dielectric (as YT pointed out) it can interfere with the high frequencies where the CPU connects to the motherboard.

 

In regard to the whole spread spectrum thing I found these products: http://www.tritech.co.uk/Products/Products-Modems.htm . The Q-PSK modulation that this thing can use looks pretty cool. With some fancy video compression, you might to be able to get decent video over this link. It looks to be quite bulky, and probably very expensive. I’d rather go with a tether.