Is a 12 volt lamp falling in a bathtub dangerous?
May 25, 2018 2:19 PM Subscribe
Please help settle an argument. Imagine a 12 volt lamp sitting next to a bathtub. The transformer for the lamp is located on the lamp cord where it is plugged into a nearby 120 volt outlet, and thus the transformer cannot get near the water without being unplugged from the outlet. The 120 volt outlet is GFCI protected. I say if the 12 volt lamp falls in the tub while still plugged into the 120 volt outlet it's not dangerous. My adversary says it could cause electrocution.
This is an interesting discussion of electrocution from what appears to be an engineering firm. Some relevant points:
- AC requires a much lower current to cause harm than DC does.
- "Sweaty or wet persons have a lower body resistance and so they can be fatally electrocuted at lower voltages."
- "Low frequency AC is more dangerous than high frequency AC."
You might be right, but I'd let you do the experiment instead of me.
posted by clawsoon at 2:40 PM on May 25, 2018 [2 favorites]
- AC requires a much lower current to cause harm than DC does.
- "Sweaty or wet persons have a lower body resistance and so they can be fatally electrocuted at lower voltages."
- "Low frequency AC is more dangerous than high frequency AC."
You might be right, but I'd let you do the experiment instead of me.
posted by clawsoon at 2:40 PM on May 25, 2018 [2 favorites]
It's the volts that jolts but it's the mils that kills.
Like, only a few milliamperes across the chest will do you in.
Current = voltage divided by resistance.
Low voltage can still generate high current.
We regularly get electrostatic shock of a few thousand volts just by brushing our hair.
posted by St. Peepsburg at 2:41 PM on May 25, 2018 [4 favorites]
Like, only a few milliamperes across the chest will do you in.
Current = voltage divided by resistance.
Low voltage can still generate high current.
We regularly get electrostatic shock of a few thousand volts just by brushing our hair.
posted by St. Peepsburg at 2:41 PM on May 25, 2018 [4 favorites]
The Navy sailor death by multimeter story is brought up often.
posted by RobotVoodooPower at 2:49 PM on May 25, 2018 [3 favorites]
posted by RobotVoodooPower at 2:49 PM on May 25, 2018 [3 favorites]
Best answer: It's not clear to me whether the GFCI would trip.
DC power bricks usually have a small capacitor between AC neutral and DC ground to reduce electronic noise added to the AC supply by the device being powered, and I don't immediately see why that couldn't be an issue with an AC to AC transformer, which could be dealt with similarly by a capacitor between the AC neutral and the AC ground; then if you provided an alternative path to ground for the current to the lamp that could result in a very small imbalance in the GFCI -- I don't know, maybe with a very sensitive GFCI?
posted by jamjam at 3:41 PM on May 25, 2018
DC power bricks usually have a small capacitor between AC neutral and DC ground to reduce electronic noise added to the AC supply by the device being powered, and I don't immediately see why that couldn't be an issue with an AC to AC transformer, which could be dealt with similarly by a capacitor between the AC neutral and the AC ground; then if you provided an alternative path to ground for the current to the lamp that could result in a very small imbalance in the GFCI -- I don't know, maybe with a very sensitive GFCI?
posted by jamjam at 3:41 PM on May 25, 2018
I'd let you do the experiment instead of me.
That's not a dare.
posted by yes I said yes I will Yes at 6:45 PM on May 25, 2018
That's not a dare.
posted by yes I said yes I will Yes at 6:45 PM on May 25, 2018
Best answer: How dangerous this could depend greatly on the design of the power supply. A conventional magnetic transformer will usually have isolated primary and secondary windings- if the secondary is floating (i.e. not referenced to ground) there really aren't any potential current paths other than between the two wires going to the lamp, so if the lamp portion falls in to the tub, any current will flow locally between the two wires and the water between them, not between a "hot wire" and any random stray grounds. (Because there is no ground reference for the secondary transformer winding.) If there was a fault or leakage between the primary and secondary windings of the transformer (very unusual failure mode) then I suppose theoretically there could be some real shock hazard, although the GFCI should then come in to play. A "switching" type power supply could be different.
posted by Larry David Syndrome at 7:10 PM on May 25, 2018 [1 favorite]
posted by Larry David Syndrome at 7:10 PM on May 25, 2018 [1 favorite]
Best answer: As far as I understand the topic: I'd say it mostly depends on the quality of the power supply, assuming it's a typical design. Whether it's an old-fashioned wall wart with a bulky, low-frequency transformer or a high-frequency switch-mode power supply, the transformer should provide isolation from the mains electrical system and the earth-ground that would create a voltage gradient with the hot side of the mains.
In several high-voltage (1000-2000 volts) isolation tests I've seen, typical switch-mode power supplies (with the little class-Y capacitor that connects the low-voltage side to the high-voltage mains side) leak less than 1 milliamp even at 2kv. At 120V (or even 240V) it's much lower. That's the only leakage from low-voltage to high-voltage, in well-designed and well-produced power supplies.
I've also seen power supplies with insufficient physical separation between high and low sides, which could cause tracking across and a bit more leakage in certain environments (like possibly humid bathrooms?) or during brief voltage spikes.
The more dangerous problem I've seen is when transformers are poorly designed to begin with, with minimal separation between high-voltage and low-voltage windings, and then built with little care to detail, wound on so tightly and/or quickly that some of the enamel insulation wears through and you get a straight short between mains and the low-voltage side. That would give you basically full contact with the mains voltage, and then you just hope the GFCI kicks in before you electrocute... too much.
The same can happen if the manufacturer cheaped out in a really unethical way and used a non-Class Y capacitor across the transformer, and it fails short, connecting mains directly with your lamp. But this - and, hopefully, the case above - would only happen if you're directly importing the cheapest power bricks from China.
Power supplies that transform mains power into low voltage are specifically designed to isolate the low-voltage side from the high-voltage, earth-referenced power grid, even when some components have failed. I'd say you'd be relatively safe (compared to unisolated mains voltage) if the 12V lamp fell in the bathtub, but you'd still want to remove it ASAP, even if just to save the lamp.
posted by WasabiFlux at 8:33 PM on May 25, 2018 [1 favorite]
In several high-voltage (1000-2000 volts) isolation tests I've seen, typical switch-mode power supplies (with the little class-Y capacitor that connects the low-voltage side to the high-voltage mains side) leak less than 1 milliamp even at 2kv. At 120V (or even 240V) it's much lower. That's the only leakage from low-voltage to high-voltage, in well-designed and well-produced power supplies.
I've also seen power supplies with insufficient physical separation between high and low sides, which could cause tracking across and a bit more leakage in certain environments (like possibly humid bathrooms?) or during brief voltage spikes.
The more dangerous problem I've seen is when transformers are poorly designed to begin with, with minimal separation between high-voltage and low-voltage windings, and then built with little care to detail, wound on so tightly and/or quickly that some of the enamel insulation wears through and you get a straight short between mains and the low-voltage side. That would give you basically full contact with the mains voltage, and then you just hope the GFCI kicks in before you electrocute... too much.
The same can happen if the manufacturer cheaped out in a really unethical way and used a non-Class Y capacitor across the transformer, and it fails short, connecting mains directly with your lamp. But this - and, hopefully, the case above - would only happen if you're directly importing the cheapest power bricks from China.
Power supplies that transform mains power into low voltage are specifically designed to isolate the low-voltage side from the high-voltage, earth-referenced power grid, even when some components have failed. I'd say you'd be relatively safe (compared to unisolated mains voltage) if the 12V lamp fell in the bathtub, but you'd still want to remove it ASAP, even if just to save the lamp.
posted by WasabiFlux at 8:33 PM on May 25, 2018 [1 favorite]
Best answer: My adversary says it could cause electrocution.
Your adversary is correct. The probability of being electrocuted is very low, but there is an unlikely combination of circumstances that could let it happen.
In order to be electrocuted, your body needs to experience a substantial current flow (tens of milliamps) through some vital part such as your heart. Such a current flow would be (I believe) impossible to generate by dumping a couple of closely spaced wires with 12V between them but no voltage with respect to anything else into the bathwater; plain water is not hugely conductive, and even if you'd loaded it up with bath salts to make it more so, the vast bulk of the current would flow in a very small region of water between and around the two wires. The results for 9V and 12V should not be terribly different in this regard, so if you want to test my reasoning, drop a 9V smoke alarm battery into your next bath and see what happens.
So any real risk would have to arise from a potential difference between the wires as a pair and something much further away from them, such as metal drain fixtures near your feet connected via metal plumbing to earth. If the wired bulb fell in somewhere north of your heart, and the bath had enough salts in it to make the water conductive, and the power supply was leaky enough to cause mains voltage to appear between its output and earth without substantial internal resistance, and there was an earth connection inside the bath - perhaps via the plughole and drain - south of your heart, and the GFCI was faulty, then you might potentially end up less lucky than Manny.
It's not clear to me whether the GFCI would trip.
If the hypothetical electrocution current were to be derived from the 12V between the two lamp wires, then I would not expect the existence of a GFCI to be at all protective; from the GFCI's point of view, all the current supplied flowing via mains active is flowing back via mains neutral and that's as things should be.
If the danger arises from a leaky power supply and an alternative return path for mains active current such as a wet body and plumbing, then a GFCI should trip before a lethal dose of current can be delivered.
posted by flabdablet at 5:13 AM on May 26, 2018 [1 favorite]
Your adversary is correct. The probability of being electrocuted is very low, but there is an unlikely combination of circumstances that could let it happen.
In order to be electrocuted, your body needs to experience a substantial current flow (tens of milliamps) through some vital part such as your heart. Such a current flow would be (I believe) impossible to generate by dumping a couple of closely spaced wires with 12V between them but no voltage with respect to anything else into the bathwater; plain water is not hugely conductive, and even if you'd loaded it up with bath salts to make it more so, the vast bulk of the current would flow in a very small region of water between and around the two wires. The results for 9V and 12V should not be terribly different in this regard, so if you want to test my reasoning, drop a 9V smoke alarm battery into your next bath and see what happens.
So any real risk would have to arise from a potential difference between the wires as a pair and something much further away from them, such as metal drain fixtures near your feet connected via metal plumbing to earth. If the wired bulb fell in somewhere north of your heart, and the bath had enough salts in it to make the water conductive, and the power supply was leaky enough to cause mains voltage to appear between its output and earth without substantial internal resistance, and there was an earth connection inside the bath - perhaps via the plughole and drain - south of your heart, and the GFCI was faulty, then you might potentially end up less lucky than Manny.
It's not clear to me whether the GFCI would trip.
If the hypothetical electrocution current were to be derived from the 12V between the two lamp wires, then I would not expect the existence of a GFCI to be at all protective; from the GFCI's point of view, all the current supplied flowing via mains active is flowing back via mains neutral and that's as things should be.
If the danger arises from a leaky power supply and an alternative return path for mains active current such as a wet body and plumbing, then a GFCI should trip before a lethal dose of current can be delivered.
posted by flabdablet at 5:13 AM on May 26, 2018 [1 favorite]
Response by poster: Thanks to all for the wonderfully technical and educational responses; hence the number of marked best answers.
posted by Jackson at 10:51 AM on May 26, 2018
posted by Jackson at 10:51 AM on May 26, 2018
The idea that an isolated supply actually improves the safety (or lack thereof) of this idea is somewhat insane to me. It doesn't matter how many transformers it's going through, that energy wants to go back to the source and has a pretty low resistance path through dirty water, your body, and the earthed plumbing fixture you're all soaking in.
Under the vast majority of conditions, that path through earth back to the grid will be high enough resistance relative to the path through the electric wires that very little current can flow, but it only takes 2mA across the heart to cause it to start fibrillating, in which case you will likely perish. Most of the time there will be more than 12k of impedance, but not if you're sitting in a bath tub and the ground is relatively moist, increasing the effectiveness of your earth bond.
posted by wierdo at 11:15 PM on May 26, 2018
Under the vast majority of conditions, that path through earth back to the grid will be high enough resistance relative to the path through the electric wires that very little current can flow, but it only takes 2mA across the heart to cause it to start fibrillating, in which case you will likely perish. Most of the time there will be more than 12k of impedance, but not if you're sitting in a bath tub and the ground is relatively moist, increasing the effectiveness of your earth bond.
posted by wierdo at 11:15 PM on May 26, 2018
Also, the idea that the mere proximity of the leads will prevent there from being a difference in potential across parts of your body is also very wrong. Electricity takes every path available to some degree or another. The fact that people and animals regularly get electrocuted by urinating on electric fences, light poles, and manhole covers that are well earthed shows the danger of that sort of thinking quite clearly.
Say the light falls in near your head. Best case, you've now got a gradient of electric potential between the source and your sewer pipe and current is flowing through your body. If you are all sealed up (no open wounds or water in contact with any mucous membranes), you'll probably be fine. If not your couple hundred ohms at best is plenty low enough to develop sufficient current to make your heart fibrillate even with relatively low voltage differences.
GFCIs aren't set to trigger at 5mA because that's a safe current, they are set that way because it is technically feasible to build reliably and inexpensively and stands a good chance of saving you from a foot in a puddle and faulty hair dryer in the hand scenario.
posted by wierdo at 11:31 PM on May 26, 2018
Say the light falls in near your head. Best case, you've now got a gradient of electric potential between the source and your sewer pipe and current is flowing through your body. If you are all sealed up (no open wounds or water in contact with any mucous membranes), you'll probably be fine. If not your couple hundred ohms at best is plenty low enough to develop sufficient current to make your heart fibrillate even with relatively low voltage differences.
GFCIs aren't set to trigger at 5mA because that's a safe current, they are set that way because it is technically feasible to build reliably and inexpensively and stands a good chance of saving you from a foot in a puddle and faulty hair dryer in the hand scenario.
posted by wierdo at 11:31 PM on May 26, 2018
The idea that an isolated supply actually improves the safety (or lack thereof) of this idea is somewhat insane to me. It doesn't matter how many transformers it's going through, that energy wants to go back to the source and has a pretty low resistance path through dirty water, your body, and the earthed plumbing fixture you're all soaking in.
The point of isolation is to increase the impedance of the path between mains active and earth via output to tens of megohms. Given a supply mains of the order of hundreds of volts, that limits the leakage current to tens of microamps. That's enough to deliver a perceptible tingle, but not to do damage.
So if the supply's isolation is working as designed, it definitely does improve safety; statistically, running this exercise with isolated supplies should kill far fewer people than doing it with non-isolated supplies.
So really, this is a question about how much confidence one has in the isolation between the supply mains and oneself for any mains-powered device in one's bathroom. Personally I prefer to rely on large air gaps (of the order of metres) for isolation between my mains-powered devices and my bathwater. That a few tens of microns of insulation inside a Y capacitor is supposed to offer sufficient isolation to prevent heart stoppage is a fact I'd rather not put to regular testing.
posted by flabdablet at 11:49 PM on May 26, 2018
The point of isolation is to increase the impedance of the path between mains active and earth via output to tens of megohms. Given a supply mains of the order of hundreds of volts, that limits the leakage current to tens of microamps. That's enough to deliver a perceptible tingle, but not to do damage.
So if the supply's isolation is working as designed, it definitely does improve safety; statistically, running this exercise with isolated supplies should kill far fewer people than doing it with non-isolated supplies.
So really, this is a question about how much confidence one has in the isolation between the supply mains and oneself for any mains-powered device in one's bathroom. Personally I prefer to rely on large air gaps (of the order of metres) for isolation between my mains-powered devices and my bathwater. That a few tens of microns of insulation inside a Y capacitor is supposed to offer sufficient isolation to prevent heart stoppage is a fact I'd rather not put to regular testing.
posted by flabdablet at 11:49 PM on May 26, 2018
The point is that it doesn't matter if the isolation transformer leaks or not in this particular application. It helps when there is a possibility a path to one of the mains conductors could exist, but not when you're leaking back to earth. There is always an electric potential between an isolated (or any other) supply and earth and even if there wasn't, a potential gradient is created in the water itself with respect to the supply when you energize it since it is not of infinitely low resistance, the exact same way a ground fault in a light pole creates a gradient in the soil.
At that point, what matters is your internal resistance and the difference in voltage across that resistance. Into and back out of your chest is only a few hundred ohms if you've been soaking, so even a couple of volts difference across that foot or two is enough to make your heart quit working if you're unlucky.
posted by wierdo at 12:41 AM on May 27, 2018
At that point, what matters is your internal resistance and the difference in voltage across that resistance. Into and back out of your chest is only a few hundred ohms if you've been soaking, so even a couple of volts difference across that foot or two is enough to make your heart quit working if you're unlucky.
posted by wierdo at 12:41 AM on May 27, 2018
Into and back out of your chest is only a few hundred ohms if you've been soaking, so even a couple of volts difference across that foot or two is enough to make your heart quit working if you're unlucky.
Sure, but according to the voltage divider principle, getting that couple of volts across a few hundred ohms by dividing down an initial supply voltage of a couple of hundred volts requires that the impedance between the mains supply and the point that's a couple of volts from earth be no more than something of the order of a hundred hundred ohms, or 10kΩ. A properly isolated supply will yield an impedance easily a hundred times that again (i.e. at least 1MΩ) between mains active and either of the output legs, with a proportional reduction in available current through the heart.
So if the isolation is actually working as designed, you're highly unlikely to die.
IF.
posted by flabdablet at 3:21 AM on May 27, 2018
Sure, but according to the voltage divider principle, getting that couple of volts across a few hundred ohms by dividing down an initial supply voltage of a couple of hundred volts requires that the impedance between the mains supply and the point that's a couple of volts from earth be no more than something of the order of a hundred hundred ohms, or 10kΩ. A properly isolated supply will yield an impedance easily a hundred times that again (i.e. at least 1MΩ) between mains active and either of the output legs, with a proportional reduction in available current through the heart.
So if the isolation is actually working as designed, you're highly unlikely to die.
IF.
posted by flabdablet at 3:21 AM on May 27, 2018
There is always an electric potential between an isolated (or any other) supply and earth
Not so. There is only a non-trivial electric potential w.r.t. earth on an isolated supply if one pole of the supply is connected to earth. A properly double-insulated isolated supply will not have a potential difference to earth that will be capable of passing more than microamps, and more likely nanoamps. Earth is not magic.
An experiment you can perform yourself. Take a battery operated flashlight - this is a 100% isolated circuit. Measure how much current you can get to flow from one pole of the battery to earth while the rest of the circuit remains insulated from earth.
So if the isolation is actually working as designed, you're highly unlikely to die.
This is correct.
O.P. Does your adversary maintain that there would be a risk with a battery operated 12V lamp?
There is zero risk of harm with a low voltage [ < 50V] battery operated device. There is a tiny risk with an isolated double insulated mains powered device - there are rare failure modes that could produce problems, but as stated above, if the isolation is working, 12V in bathwater is harmless.
posted by HiroProtagonist at 8:05 PM on May 27, 2018 [1 favorite]
Not so. There is only a non-trivial electric potential w.r.t. earth on an isolated supply if one pole of the supply is connected to earth. A properly double-insulated isolated supply will not have a potential difference to earth that will be capable of passing more than microamps, and more likely nanoamps. Earth is not magic.
An experiment you can perform yourself. Take a battery operated flashlight - this is a 100% isolated circuit. Measure how much current you can get to flow from one pole of the battery to earth while the rest of the circuit remains insulated from earth.
So if the isolation is actually working as designed, you're highly unlikely to die.
This is correct.
O.P. Does your adversary maintain that there would be a risk with a battery operated 12V lamp?
There is zero risk of harm with a low voltage [ < 50V] battery operated device. There is a tiny risk with an isolated double insulated mains powered device - there are rare failure modes that could produce problems, but as stated above, if the isolation is working, 12V in bathwater is harmless.
posted by HiroProtagonist at 8:05 PM on May 27, 2018 [1 favorite]
This thread is closed to new comments.
posted by Dr. Twist at 2:32 PM on May 25, 2018 [1 favorite]