3
\$\begingroup\$

Tradition says that electricity and water do not mix well, but sooner or later some electronic components are going to come into contact with water. I noticed that datasheets don't mention anything about that for regular components so I'm looking for a source of information on that.

For example I noticed that some resistors with copper tin plated leads will release tin and copper when submerged in water and powered on (looks like electrolysis is taking place). I also noticed that some electric motors work just fine when submerged in water and that for example AA alkaline batteries don't seem to have any problems working under water. I also heard that some types of mineral oils can remove insulation from cables and that some people who used mineral oils to cool computers had problems with that.

Books are acceptable, but bonus points if the source if freely available on the Internet.

\$\endgroup\$
  • \$\begingroup\$ Some cables are specified to withstand oil (for example Lapp Ölflex). \$\endgroup\$ – starblue May 3 '11 at 22:02
  • \$\begingroup\$ What is the actual problem you're trying to address? \$\endgroup\$ – Nick T May 3 '11 at 22:27
  • 2
    \$\begingroup\$ @Nick T There's no particular problem at the moment. I was hoping to get some information on failure modes and as you said it all sorts of problems with high-speed and/or high-impedance circuits. I did have some simple projects which should be able to survive rain in mind, but I was planning to use urethan based coating on them. \$\endgroup\$ – AndrejaKo May 3 '11 at 22:37
  • 2
    \$\begingroup\$ @Kortuk♦ What's DI water? Deionized maybe? \$\endgroup\$ – AndrejaKo May 4 '11 at 7:00
  • 1
    \$\begingroup\$ @Andrejako, Sorry, I forget that at one point I had to learn the slang also. We reference it regularly, and yes, it is deionized to 10 Mohm. And yes, when you expose the water to air it might as well be normal water in under an hour(in my experience in under 5 minutes). If you do not know what a megger is already, look into it. It will be key to checking insulation. \$\endgroup\$ – Kortuk May 4 '11 at 9:32
9
\$\begingroup\$

The easy answer is, things don't work underwater. Although technically water is not conductive (a circuit will run fine when submerged in pure water), it's all the impurities in the water that make bad things happen. And water will pick up lots of impurities. Take a container of pure water and just let it sit there, uncovered. Over time it will begin to absorb stuff from the air and its electrical conductivity will go up.

To make matters worse, water is a mild solvent so things tend to dissolve in it. And there's your real problem. You cannot predict what will be dissolved in the water. Is the pH acidic, or alkaline? What else is on your PCB that could get dissolved in the water that then effects other things? Maybe the ink on the labels of those caps could turn the pH slightly acidic, causing more corrosion on something else. Since you don't know what's in the water, you can't predict what will happen-- but only bad things can happen.

It's unlikely that you'll ever see an electrical component that is spec'd for operation underwater. The only spec that I've ever seen is in the temperature spec of some chips, that's listed as "+120 to -40 deg C, non-condensing". Underwater is considered "condensing", by the way. :)

The normal way of protecting electronics from water or high humidity is to use a conformal coating on them. That's basically a way to cover the PCB with a protective coating.

There are people who cool PCB's by submerging them in some sort of liquid. Normally mineral oil or some sort of inert chemical (the name escapes me right now). I have only seen anecdotal evidence for this, although maybe the supercomputer guys from 10-20 years ago would have done a study.

And somewhat off topic, but... While a standard motor might work underwater, it wouldn't work for long. The water, and motion of the motor, would tend to remove the lubrication. Followed by corrosion. Then the thing would seize up.

\$\endgroup\$
  • 2
    \$\begingroup\$ Regarding pure water - your water may be pure now, but after soaking your PCB, it won't be pure any more. The inert chemical you may have been thinking of is Fluorinert. \$\endgroup\$ – W5VO May 3 '11 at 21:29
  • \$\begingroup\$ Flourinert (en.wikipedia.org/wiki/Fluorinert) \$\endgroup\$ – Toybuilder May 3 '11 at 21:55
  • \$\begingroup\$ conformal coating guards against humidity, but for serious immersion you're looking at potting. \$\endgroup\$ – JustJeff May 3 '11 at 22:21
  • \$\begingroup\$ @JustJeff, based on my research conformal coatings often gaurd against water better then humidity. At my last job we were having a problem where the humidity, which can move through the conformal coating easily, would condense on the board under the coating and cook components. \$\endgroup\$ – Kortuk May 4 '11 at 0:59
  • \$\begingroup\$ @Kortuk - well, the liquids thing is a whole spectrum of problems. there's humidity, there's rain, there's immersion, pH, salinity, pressure, all that. my point was just that most things I've seen that are intended for submersion are potted in epoxy (or something like it) \$\endgroup\$ – JustJeff May 4 '11 at 1:32
5
\$\begingroup\$

Unless a part is explicitly designed to work underwater, a manufacturer would just laugh in your face if you asked them about it. Sure, they may work for a time, but you would need to test and certify them yourself. The manufacturer's specifications and guarantees will not apply if you use them outside of their normal conditions, plus the water would give you all sorts of problems with high-speed and/or high-impedance circuits before it's done eating through traces.

The only environmental conditions electronic components are expected to survive are temperature swings (and occasionally, radiation). Water, given enough time, is extremely corrosive and will destroy unprotected, exposed circuitry. It is the job of enclosures and coatings to protect parts and PCBs from moisture and other potentially corrosive chemicals.

An enclosure would be best if you want to survive underwater for mechanical reasons, but if you're trying to cut down on space/weight, you may be able to get away with just a conformal coating—essentially a thin layer of potting. A company I used to work for would get warranty returns for vehicle data computers that were half filled with water but still fine electrically as the board was conformal coated with a silicone resin.

\$\endgroup\$
4
\$\begingroup\$

The prudent answer to insuring that your project will survive in a harsh environment (rain, underwater, whatever) is to properly protect it. You can put it in a waterproof box or potting it in epoxy will do an excellent job of protecting it.

Of greater concern perhaps is safety, operating any sort of energized electrical/electronic device in the open (rain or not) where someone could accidentally touch it is not a good idea. Would you want to step into a bathtub that had a live electrical wire in it?

\$\endgroup\$
2
\$\begingroup\$

Large electronic valves used in radio transmitters often had their anodes cooled by direct contact with de-ionised water. As long as the water was pure the insulation was sufficient for the several 10's of kV on the anodes. Interesting if the pipes burst as the anodes would overheat very quickly.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.