4
\$\begingroup\$

I have a board that is mounted inside a waterproof housing, but occasionally the housings develop leaks and expose the internal circuits to water.

This can cause unpredictable failures as the corrosion slowly grows until a short or open develops someplace on the board.

I am wondering if there is such a thing as a "liquid fuse"- that is, a component that will permanently fail open-circuit upon exposure to liquid. I'd put one of these right where the power enters the board so it would cut off current to the rest of the board before corrosion could cause a short someplace.

I imagine the component itself could be some kind of conducive powder held together with an easily dissolvable binding agent. Or maybe a jumper compressed by a spring, and the spring is held by some dissolvable material?

This would be going into a mass produced product, so the part would ideally be SMD reflowable, cheap, and commercially available.

I've looked for a part like this, but can't find anything. Any ideas?

\$\endgroup\$
  • \$\begingroup\$ The method used by manufacturers such as Apple is to use a permanent visual indicator that changes color on exposure to water. That voids the warranty so eliminates the problem (for them). \$\endgroup\$ – Spehro Pefhany Jun 6 '15 at 23:44
  • \$\begingroup\$ Absolutely, but for this issue I don't care about having a witness to the liquid intrusion (trust me, you can tell when you get one of these back :) ), I am just looking for a way to fail fast and predictably rather than letting a short eventually happen someplace where it could cause unpredictable and bad behavior. \$\endgroup\$ – bigjosh Jun 7 '15 at 0:00
  • \$\begingroup\$ The old VCRs had condensation detectors (an interdigitated pattern) that would be used in a circuit to remove power from most of the circuitry. \$\endgroup\$ – Spehro Pefhany Jun 7 '15 at 0:02
4
\$\begingroup\$

EDIT: What I forgot about with the lithium option is that it also reacts with Oxygen and Nitrogen, but it was a fun thought while it lasted. Thanks to Dmitry for the comment.

If you have access to a manufacturer that can do some tricks with metals, you could do a trial run of a small resistor-type component with an exposed lithium link.

Lithium reacts with water, but not quick or violently enough to melt or explode. I have no idea how "legal" that would be though, because it will outgas. And well, lithium is scary now.

And you will have to "dessicate" the box before revealing the lithium, or it might start oxidising with the existing moisture level in the air present.

Otherwise you could make a leakage sensor strip that triggers a destructive process with a MOSFET or some such.

schematic

simulate this circuit – Schematic created using CircuitLab

If the leakage between the traces (make them without Soldermask and possibly ENIG?) starts beating the 22M resistor it will start shorting out the power, draining a lot through the fuse, it will "vaporise" soon enough.

It may be needed to make it a two-stage system, because this can take a while to start conducting through the MOSFET and you don't want it to break, because that is your safety, like this:

schematic

simulate this circuit

In this case the MOSFET will only need to start conducting a few mA for the PNP transistor to boost that up nice and strong, making the reaction much quicker.

You can also put a P-MOST in place of the PNP, this will really boost up the reaction, but it might also become "too hair-trigger". If you make the pinch-off resistor (R3) even larger with the P-MOST it might start triggering with tiny little spikes of leakage that you might still want to ignore.

But you have to realise that this will always only "sense" where you put it on the board. The advantage of this is, with a fuse on the input you can riddle the board with shorting-mosfets that will blow the fuse. If the current consumption of your device is low, you can keep the fuse small and you could use SOT323 type tiny transistors, wasting little space.

You will need to tune the component values (and types: PNP/P-MOST) during some tests, you can estimate the initial values for the 22M resistor when you know the distance between the interlocking traces and what kind of water creeps in. I'd advise you to make the interlocking traces the once with the least clearance on the board, that gives you some safety: They'll be the first to start leaking noticeably.

\$\endgroup\$
  • \$\begingroup\$ I had thought about connecting the on-board MCU to a sensor grid as you suggest and then having it issue the self destruct by shorting out a standard fuse, but your way is more fail-safe with minimal additional cost. Still, I feel like this must a common problem and there must be a standard single component physical solution, no? Thanks! \$\endgroup\$ – bigjosh Jun 6 '15 at 23:48
  • \$\begingroup\$ Not really, firstly because 99% of the market couldn't care one bit about water, they'll just say "Oh, well, warranty void, buy a new one, your problem". And the other 1% would employ shrink-free casting materials or improved housings to guarantee no moisture could touch the PCB during its expected life cycle. But if I wanted to do something like this, I think I'd go for my option two. But, as a hint: I do stock 40liters of shrink-free casting resins. \$\endgroup\$ – Asmyldof Jun 6 '15 at 23:51
  • \$\begingroup\$ :) I've looked into adding a conformal coating the boards as one option, but a fail safe fuse might be better since beside the board there are other things inside the casing that can not be make water-proof (like a motor), so I'd prefer to kill all power before corrosion starts. \$\endgroup\$ – bigjosh Jun 6 '15 at 23:57
  • 2
    \$\begingroup\$ Lithium (or sodium, calcium etc) are not a good option, because they will react with the oxygen. You will need to make cases as airtight as light bulbs - rubber joints will stop water but not gas. And you won't be able to ever open your devices for repairs. \$\endgroup\$ – Dmitry Grigoryev Jun 7 '15 at 9:06
  • \$\begingroup\$ @DmitryGrigoryev Forgot about that little detail. In fact, filling it with nitrogen would be as bad, since it also makes nitrides happily. I'll make an edit for scientific accuracy. \$\endgroup\$ – Asmyldof Jun 7 '15 at 13:35
2
\$\begingroup\$

In general I agree with EM Fields that the solution is a better enclosure.

The trouble with a component that detects water drops (be it a pair of conductive tracks or an aspirin/sugar pill holding a switch closed) is that the water will inevitably miss the sensor and short/corrode some other part of the board.

Therefore, what you have to detect is atmospheric humidity inside the enclosure. You will have to include some dessicant in your device to ensure the humidity remains low.

Here's a list of dessicants. http://en.wikipedia.org/wiki/List_of_desiccants Unfortunately some of the really strong ones are missing so see also here https://chemistry.stackexchange.com/a/26378/5541

One possible way of making a humidity sensor is to use a hygroscopic substance. Dilation of nylon thread was once used in airconditioning systems. http://en.wikipedia.org/wiki/Hygroscopy

Humidity sensors are available (probably now on different principles) but it still leaves the problem that the moisture, dessicant and sensor will be in different locations.

An alternative would be to use a pill made of deliquescent substance to hold a switch closed. That is, one that absorbs humidity from the atmosphere (like a dessicant) and as a consequence dissolves in it (like the aspirins and sugar pills mentioned.) This would be more reliable than aspirin or sugar. The wikipedia page on hygroscopy has a link to a video showing deliquescence of Calcium Chloride, but a stronger dessiciant like Phosphorus Pentoxide might be more reliable. The reaction is quite slow (due to the poor mass transfer of water from the air to the locality of the calcium chloride) and I believe the video has been sped up. Note that a mechanical switch may not be absolutely essential, as the solution, if kept in place, would be conductive (in contrast to the solid which is an insulator) and therefore could be contained in some kind of wadding to detect a conductivity path.

Conductive paths made of sodium or lithium would fail open circuit on exposure to moisture, but also more slowly on exposure to oxygen. Also I think you would need a film rather than a wire, as i think the wire would respond too slowly.


If you really need to detect a breach of containment quicky, I think the best thing to do is what the food industry does: use vacuum/pressure. We've all seen vacuum packed foods with a button on the lid that says "reject if button can be depressed." If you have a suitably strong enclosure, all you need is a pressure sensor and a means of evacuating the enclosure.

The good thing about vacuum is that (when it is a strong vacuum) it is a steady pressure reference, but if you want to avoid things leaking in and prefer to have them leaking out, you should use pressure instead of vacuum. Bear in mind though, that the temperature will affect the pressure quite a lot. If you seal your box at 0C (273K) with an internal pressure that's barely detectable, by the time it heats up to 27.3C you will have a pressure of 1/10 atmosphere inside, about 1.5psi or 10kgf/cm2. Note that this system is used succesfully in central heating boilers to check the containment of water in the heating circuit through the radiators (which is isolated from the supply and consumer taps for a variety of reasons.) These circuits are usually pressurised to about 1-3bar.

In potentially explosive atmospheres (such as may occur during gas leaks in oil fields), the European AtEx regulations permit instrument cabinets to be pressurized continuously with a supply of compressed air as one of various possible forms of protection. This obviously ensures that no explosive atmosphere (or other undesired substance such as humidity) can enter the cabinet.

\$\endgroup\$
  • \$\begingroup\$ I think positive pressure rather than vacuum would be the better option, firstly because it will to some extent prevent water ingress rather than encouraging it, and secondly because most electronic components are not meant to operate in vacuum and have unpredictable outgassing characteristics that will degrade the vacuum over time. The other thing is that relative humidity sensors will work more sensibly if placed in an atmosphere that is not 100% water vapor. \$\endgroup\$ – Oleksandr R. Jun 7 '15 at 13:41
1
\$\begingroup\$

If you're planning in making and selling a zillion widgets, any band-aids that you can think of up front to try to ameliorate the damage after the fact; that is, once the watertight integrity of the enclosure has been compromised, will do nothing but bring you grief and sleepless nights.

What you should do, in my opinion, is to spend the time and the money needed now to make sure the housing CAN'T leak, and not worry having to run around like a crazy person, later, trying to put out fires that should never have started.

But that's just me, and you may have an entirely different, and valid, reason for going about it the way you are.

But, if all else fails and you do get a leak, it should be fairly easy to spring-load an aspirin and allow a mechanical switch to open/close when it [the aspirin] dissolves away.

\$\endgroup\$
  • \$\begingroup\$ Agreed, but for this application I do think it would be appropriate to add an extra layer of protection. You can't prevent all leaks in a consumer product, you should see how some people treat a product once they get it home. :) I love the spring loaded aspirin switch - this is exactly what I am looking for. Have you ever seen anything like this available as a commercial product? Thanks! \$\endgroup\$ – bigjosh Jun 6 '15 at 23:55
  • \$\begingroup\$ I looked around a bit on Google, but came up empty-handed. I'm not so sure that aspirin's a good idea, it mighty introduce corrosion problems of its own. Something like a sugar cube should be OK, though; fast dissolving and anionic. Hmmm... maybe there's a product there? \$\endgroup\$ – EM Fields Jun 7 '15 at 0:20
  • \$\begingroup\$ Are you sure your customers won't expose the box to shocks intense enough to trigger mechanical water protection system without any water? \$\endgroup\$ – Dmitry Grigoryev Jun 7 '15 at 9:01
  • \$\begingroup\$ For this application I'd be happy to have some false positives due to mechanical shock. Thx. \$\endgroup\$ – bigjosh Jun 8 '15 at 1:06
0
\$\begingroup\$

To unabashedly stea^H^H borrow an idea from @EMFields, you could mount two right-angle normally open tactile switches facing each other and separate them with a sugar pill or similar.

enter image description here

Mill slots under the switches in the PCB so that corrosion will be less likely across the switch.

Aspirin is a drug so it might (paradoxically) actually cause headaches in this case, but then sugar might attract vermin (ants, or fruit flies, for example).

enter image description here

Another possibility would be to have two close-spaced serpentine patterns of very thin conductors on a replaceable PCB (like the differential pairs that your PCB package can probably generate automatically). Remove the solder mask from those traces and put them where they will see the water first. Power goes through those traces. The corrosion will preferentially take place in that area and will eat through the traces, cutting the power. Simple and crude.

enter image description here

\$\endgroup\$
  • \$\begingroup\$ +1 for the first ever fruit fly to grace an EE.SE answer. You could also use something less edible like salt. \$\endgroup\$ – tomnexus Jun 7 '15 at 21:10
  • \$\begingroup\$ @tomnexus salt or maybe vitamin C citric acid sure. Had a young fellow that brought a fruit fly infestation down on us, so I sent a giant version of that photo to everyone by email. Can't be leaving banana peels around... \$\endgroup\$ – Spehro Pefhany Jun 7 '15 at 23:35
0
\$\begingroup\$

Make a printed circuit board with dozens of fine pitch parallel traces to carry power. In a separate area of the board an op amp or mosfet circuit detects conduction between traces not covered by conformal coating and ceates a power and ground short. Thin traces on the pcb burn up and power is removed. Repair replaces the "sensor/fuse" board and leaky enclosure.

\$\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.