Everybody knows that water conducts electricity. Almost everyone on this board probably knows it is because the water contains dissolved ions that carry the charge, and that distilled water is an insulator.
So, long story short, I had the brilliant idea of cleaning off my TI-calculator’s battery contacts with water. I mistakenly thought there were no electronics in the area the water was being poured on, and was quite wrong. Turns out, the main circuit board is right behind the battery compartment on the Nspire CAS. It has a couple of relatively sophisticated BGA parts on it, probably a system on a chip and a flash or a RAM but I didn’t look up the chips.
Anyways, why did my calculator not run when wet? Those are all digital parts there, and the water is not very conductive, and 3.3 volts is not very high. Why wasn’t the extra little bit of voltage bridging, say, a logic 1 on a pin and a logic 0 on an adjacent pin, too low to matter?
I managed to get the cover open and clean it off with alcohol, then I let it dry out overnight and it of course works fine now.
For just the pins on on chips water is a good enough conductor to cause a short that will result in a signal being misread. Not every ‘high’ signal on a pin is exactly a 3.3 volts, or every ‘low’ input exactly at 0 volts, and they must maintain voltage during a critical time frame to be read correctly. It doesn’t require a huge change to the waveform you would see at a pin to interfere with the logical operation of the chip, and that doesn’t even get into the potential physical damage that can be done.
IANAEE but flooding a circuit with water just shorts the hell out of everything, not just a pin here and a pin there. The magnitude of the voltage doesn’t matter. 3.3v is enough to drive the circuit so it’s high enough. I suppose you’re lucky it still works–that kind of shorting can cause current spikes that could fry certain components.
Your question starts with “Why” and someone else may need to provide a more detail technical answer.
BTW the textbook method for cleaning contacts is to rub with a pencil eraser.
FWIW I needed to control a electric motor’s speed using a volume control for a piece of audio equipment. I knew I would far exceed the power that was normal for the audio equipment and would most likely fry it. So I put the volume controller under water and it worked, I could actually see bubbles from boiling water coming out of the controller, but it worked just fine for what I needed to do. Shortly after I ran it without the water and the controller smoked and melted and stopped working pretty soon as power was supplied.
Just speculating here, but I wonder what the impedance of a wet I/O is. Perhaps it screws up the transmission line characteristics enough to make the part read the wrong value of the signal. That’s less damaging than a short, but still causes a fail.
Was this a potentiometer? Just a simple variable resistor? Because yeah, you’d expect it to work fine under water, the water is not really a short, it’s just a voltage leak from the high to the low side of the resistor, and equivalent to sticking a big resistor in parallel to the variable resistor.
There’s that and any effects on the conditioning of the lines such as shorting out a resistor or capacitor.
This does make me wonder if there’s some kind of sealant that could be sprayed on electronics that people often get wet such as cell phones. Obviously the whole thing could be encased in plastic, but I’m think of a simpler less costly spray on material.
And…I think I know why my calculator failed. It’s none of the answers above.
It’s simple - I suspect the problem is the water is not as good of a resistor as the standard 10k ohm pullup/pulldown resistors that most electronic circuits have a dozen of. So circuits that the designer intended to have a high or a low voltage state when idle may end up in the wrong state. This is why it didn’t even power up - to even power up a circuit like this, you do things like pull down the reset pin to ground usually when it is running.
It’s just a theory, I would have to confirm it with pretty time consuming experimentation, but I think that’s the explanation.
Though, Tripolar, sticking an extra 8k ohm resistor - or whatever the effective resistance of tap water across a gap of a millimeter actually is - randomly around could discharge capacitors, messing up RC filters, and so forth. But a simple all-digital circuit like this has nothing like that. The main components are just traces, pullup/pulldown resistors, and these caps between power and ground called “decoupling capacitors” that act to filter the power. Shorting a decoupling cap shouldn’t do anything - the power might drop a little, from 3.3 volts to 3.25 or something, but it should still work.
Shorting a pullup/pulldown resistor, however, would, because there’s little energy flowing through the pulldown/pullup to restore the state you are affecting with the bridging with the water. (this is why they are used, you can change the state of a node being sent to a default state with the pulldown/pullup very easily)
De-ionized water is actually not a very good conductor: 10MΩ/cm.
But you were washing out battery electrolytes. The remaining solution is, by definition, a weak electrolyte solution and therefore pretty conductive. The water left in the circuit probably had notable concentrations of metal salts and other conductive ions.
That would be worse than tap water, let alone distilled or purified water.
Huh. Yeah, I didn’t use distilled water, either. See, I thought the compartment in the bottom of the calculator was just a hollow holder for the battery, and that the actual main circuit board was up by the LCD display. Hefting it, it feels like that, and when I took it apart, I think they could have built it that way. But they didn’t…
Anyways, you hear stories of people dropping their iphones into the sink and they also quit. What gives, there?
Except that the drop might be from something less than 3.3 volts to start with and that becomes another factor to combine with changes in impedance. In addition the shorts could be brief causing small voltage spikes and dips that throw off the internal logic.
I don’t think the specifics for this one case are as important as not getting your electronics wet in the first place.
In the short term (pun intended), water creates a low-resistance path which messes with the inputs to all the electronics. Even if you start with distilled water, there is all kinds of gunk in an electronic device (dried sweat, for example), which means that the water creates shorts for high-impedance circuits.
That effect is usually not fatal - if the device is dried out quickly enough, it will work. But, what happens to phones is this: the battery (which is hard to remove in an iPhone) causes the water to corrode any electronics it touches, and often completely etches through conductors. If the battery can be removed fast enough, the phone will survive. If not, it’s generally ruined.
Expanding on the question - why aren’t modern electronics made invulnerable to water? Or at least less vulnerable? Could the circuit boards in a phone be coated with something waterproof like epoxy, save for the connections to speaker, display, etc?
Or, for a device like an iPhone that is effectively a sealed thing with no user-serviceable parts, why not actually seal it? Not looking for waterproof to 50 meters, just something that won’t be destroyed when someone knocks over a nearby drink.
I asked the same question. They can be encased in plastic, but that adds weight and cost to portable devices. A light coating of epoxy should do much better than nothing at all. GoPro manages to seal their cameras, I don’t know why the same can’t be done for a cell phone.
It would add extra weight, cost, and size, all of which are bad things for a cell phone. You also have to be careful with heat dissipation when you start sealing things up.
But it can be done, and in fact there are quite a few cell phones out there that are water resistant, if you want to pay the extra money for them. They aren’t quite water-proof and you probably wouldn’t want to dunk one in your bathtub, but they are designed to handle someone knocking their drinking glass over and having it spill all over the phone.
Phones can be made water-resistant (Samsung makes them, among others). But, it clearly increases the cost - not only do all the case openings need to be sealed, but all the switches need to be waterproof, as well as all the connectors (headphone, charger, etc.). The speaker and Mic also have to be waterproof. Apple apparently doesn’t think it’s worthwhile.
That’s going to be part of it. Tap water has a resistance of something like 20 ohm-m. Suppose you have a signal pin with a pullup right next to a ground, and a drop of water shorts the two pins. The pins might be quite close together; say 0.25 mm. And the cross section in the ballpark of 2x1 mm (accounting for the fact that the droplet bulges out somewhat). That’s 2500 ohms, which is enough to overwhelm that 10k pullup (which might actually be 100k or more). At 3.3v, that’s a >1 mA current, which is enormous.
Lots of signal pins aren’t going to be happy at all about sinking 1 mA to ground; this is one reason why you can easily fry electronics if you power them when wet.