Why we ground electrical systems (2015) | Hacker News Reader

Why we ground electrical systems (2015)

reddit.com

119 pointsterminalhealth117 comments

I'm actually the person who wrote this Reddit post. I'd never heard of Hacker News before, but someone messaged me on Reddit telling me this 4 year old post just surfaced over here.

Let me know if you have any questions about electrical theory or installation, or anything else!

Question with myth #1 that "electricity is trying to get to ground". Would that be an accurate statement if you replaced "ground" with a place of lower electrical potential?

Or can you elaborate but the meaning that it's trying to get back to it's source?

Well, electrical potential is only defined as a relation between two points. So the neutral point on a transformer, for example, doesn't have any electrical potential in and of itself. It has electrical potential with relation to, say, the termination point of one of the phases.

The electrical potential between any two given points is dependent on the impedance along the pathway between the two. The lower the impedance, the lower the potential (or work/energy needed to move current). So, yes, in a way electricity wants to get to the point with the lowest potential, but that point will always be the point where the electricity originates.

I was a little vague with "where the electricity originates" because that could be a number of things. It could be a battery, a transformer, a generator, a turbine, a dynamo, etc. Whatever is creating the difference in potential between two points is the "source" of the electricity.

I've always preferred to think of it as pumping water from the bottom of a (practically infinite) reservoir and dumping it at the top of a mountain then doing some work as it flows downhill, back to the reservoir.

It provides a nice visual for why current always makes it back to ground, just like water always flow downhill. It also removes the tendency to anthropomorphize electrical current and say things like it "seeks out" ground. When something falls from the sky we don't say it's trying to find the ground!

Yes water is a great way to describe electricity in a lot of ways. I'm just struggling with the "returning to where it originates". For instance the electricity didn't originate from the grounding rod it came from the transformer. Earth is providing a very low electrical potential that the electricity is "attracted" to or "falling" to in the reservoir example. Am I wrong in thinking that all electricity is flowing back into earth?

> Am I wrong in thinking that all electricity is flowing back into earth?

Yes, sadly, you are wrong. But it's not a strange error.

Electricity always flows in a circle(circuit).

And, in fact, in medical devices, transformers are often used to completely isolate devices from the line that they are plugged into. Electrons on the device side of the transformer will NOT try to flow back into the line side or an earth. They only want to flow back to the device side of the transformer.

Now, the issue is that when you want to create really strict isolation like this, suddenly all manner of things that normally you don't pay attention to suddenly become relevant. Is the device side of that transformer really not connected to the line side anywhere? No goop on the board? No water vapor? No lines that are a little too close? Is the hospital bed not connected to anything?

Guitar players who use tube amps and vocalists who use condenser microphones wind up with this issue all the time. Both the amps and the mics are "isolated" with relatively high voltage signals floating around--300-400V for amps:48V for mics--and consequently strange paths cause lots of "buzz" in the signal.

Yes instinctivly it wasn't making sense as I knew the circuit needs to be "completed". So if that's the case then what is the point of the ground rod that the neutral bar/wires in your breaker box are connected to?

This article posted in a comment above explains it well: http://amasci.com/amateur/whygnd.html

Essentially, the ground rod acts as an "anchor" holding the neutral wire and the ground at the same electric potential. If there was no ground rod, the earth and the circuit would be "floating" relative to one another, and a dangerously large voltage could develop between them.

When it comes to AC (even discontinuous DC), closed loops are not necessary for current flow. A basic example is a capacitor, which can permit current while it builds up an electric field between is plates. A related example is an antenna, where AC radiates energy that can vibrate electrons in a far away antenna. (Radio telescopes are proof an antenna doesn't need a ground or a closed circuit to receive.)

Closed loops ("circuits") are part of the paradigm, so yes they are necessary at the level of abstraction we generally analyze electronics in. Sure if you came from Pluto with a metal sphere teeming with excess electrons, they would disburse themselves on our planet and never return home, but that isn't circuit analysis!

For example, the current into one terminal of a capacitor does equal the current out of the other terminal. Yes electrons are building up on one plate, but they're being depleted from the other - the net charge of the capacitor remains zero. If this weren't true, differently-charged capacitors would be physically attracted to one another!

If this goes against your intuition, it's because you've become so accustomed to this abstraction of ground which (mostly) lets you forget about the current flow on the other leg of the capacitor.

(In the electronics realm, the use of the term "ground" is much more casual compared to the NEC. To put it in terms of OP, "ground" in the electronics realm is more akin to the "grounded conductor", but alas could actually refer to anything you feel like thinking of as the reference.)

I only brought up ground (in the Earth ground sense) because I've heard people think that antennas in general work by using the Earth as a second conductor, with EM radiation as the first. Antennas are conductors that are part of an un-closed electrical network, relying on the capacity of a conductor to hold an oscillating non-equilibrium charge (though they can be modeled as an RLC network). My previous comment was written out of a reasonably common conception of a closed circuit being a loop with constant current flow, but capacitors break such a circuit. (I don't mean to imply what bsder said was in any way wrong, just not necessarily the whole story depending on how it's interpreted.)

The realm of electrodynamics is pretty interesting, and it's where seemingly basic concepts like electric potential begin to fail -- it becomes path-dependent!

> The realm of electrodynamics is pretty interesting, and it's where seemingly basic concepts like electric potential begin to fail -- it becomes path-dependent!

Not really. The problem is that we have sort of an "electron abstraction" which is incorrect in the Heaviside-Hertz pedagogy when fields start to store energy. I recommend "Collective Electrodynamics" by Carver Mead as a modern formulation without the silliness of an Aether: https://www.amazon.com/Collective-Electrodynamics-Quantum-Fo...

> I only brought up ground (in the Earth ground sense) because I've heard people think that antennas in general work by using the Earth as a second conductor, with EM radiation as the first.

Really? Most of the time I describe antennas as sort of really long range transformers. And, while you need each side of the transformer to be a circuit, the two sides of the transformer don't have to interact other than through a field.

> My previous comment was written out of a reasonably common conception of a closed circuit being a loop with constant current flow, but capacitors break such a circuit.

Yes and no. Capacitors have the hand wavy notion of "displacement current" in classical electrodynamics--but most of the issue is with the fact that we use the Heaviside-Hertz pedagogy which was formulated back when everybody believed in the Aether.

The real issue is that to deal with capacitors you must deal with fields rather than just the notion of electrons. However, if we kind of squint and wave our hands the "electron formulation" can be kinda sorta made to work. (Side note: Capacitive dividers are even more annoying and you have to be really careful.)

Classical Heaviside-Hertz electrodynamics also has a lot of issues dealing with motors and generators, as well. Again, the key is that an "electron formulation" isn't really enough when fields start holding an appreciable amount of energy.

Thanks for the pointer, and I look forward to checking out that book.

(Re "not really": with Maxwell's equations, the curl of the electric field is generally non-zero, so scalar potentials are not well-defined, which is all I meant. My experience is having gone through Purcell long ago, and by now I have some familiarity with [mathematical] gauge theory, but really I'm only an E&M dilettante.)

When you talk about "electricity" you need to distinguish between "current" and "charge carrier." Current is the amount of charge that flows through a point per second. The charge carriers are the electrically charged particles that actually move. Where current goes is a matter of what we're trying to do with a circuit, where charge carriers go is why we need ground.

What a voltage does is pull or push the charge carriers. When lots of them flow, you have a current. Conductors, like a wire or ground rod, are full of free electrons to act like charge carriers (kind of like a pipe filled with water, the voltage is a pump that moves it).

The duality of pulling/pushing charge carriers is why we need a circuit. In order to push charge carriers, we need something to pull them from (a source) and somewhere to dump them (a sink). When we have no source and no sink, charge carriers have nowhere to come from and nowhere to go.

Ground is a convenient source/sink for charge carriers because it's roughly uniform in charge and huge, so pulling tons of charge carriers from it doesn't impact it greatly.

And it's not that charge carriers are always flowing back to earth, but back to their source. That's why ground is sometimes called a "return path." To move a charge carrier, you need to give it potential. It will lose that potential and return to the point of lowest potential difference from its origin - which is its origin.

But that said, for things like AC power, the charge carriers aren't actually moving very far at all and have a net displacement of 0. They vibrate adjacent charge carriers, and we convert that vibration into unidirectional (DC) voltages that can push/pull from local sources/sinks, be it the literal earth (mostly for safety ground) or a small plane of copper on a PCB.

Hmm okay starting to make more sense now. Do you know a good visual explanation that shows the "path" electricity takes from the transformer to your home and back out to the ground rod. Note: I fully understand in AC the electrons aren't actually moving along this path. But I guess I don't see how the circuit is ever "complete" or a circle.

It doesn't go "back out to the ground rod," it goes back to the transformer. Note this paragraph in the parent comment:

> And it's not that charge carriers are always flowing back to earth, but back to their source. That's why ground is sometimes called a "return path." To move a charge carrier, you need to give it potential. It will lose that potential and return to the point of lowest potential difference from its origin - which is its origin.

This video has a great visual explanation of grounding, imho.

https://www.youtube.com/watch?v=P-W42tk-fWc

This was exactly what I was looking for thank you!

Electric potential is created by a difference in electric charge between two points. Because it's a measurement between two points, it is never absolute -- "a very low electric potential" is meaningless unless you've defined what it's low relative to. A battery/electric outlet/transformer creates a potential difference across its terminals, and that's why electricity wants to flow from one terminal to the other. Usually, the terminal with the lower (more negative) potential is defined to be the 0V measurement, but this is arbitrary and out of mathematical convenience -- we could call it 1V or -1000V, and the circuit would still work the same (all that matters is the difference between the two terminals, not the absolute value of either terminal).

The neutral lines of circuits are often tied into Earth, making the voltages of the neutral line and the earth equal to one another at what we've defined to be 0V (for reasons of convenience and safety). There's no intrinsic property of Earth that gives it a low electric potential, and electricity doesn't intrinsically want to flow back into Earth.

This analogy doesn't work so well for AC, which involves both pushing and pulling on electric charges.

It's more like pipes full of a gas like air. The power plant has a big reciprocating piston that is pushing and pulling on the gas, creating a pressure wave. One side of the cylinder is "aired," meaning it is in contact with atmospheric air. These pipes make their way to clients, who attach the pipes to equipment of their own, for example a piston that converts this wave back into mechanical energy. Again, one side of this piston is in contact with atmospheric air, which is the reference pressure for the piston.

Sometimes the pipe develops small holes, and if a hapless worker gets too close, they can either get cut from an out-blast or hurt from smashing against the equipment when the air is sucking in (both being from the difference in the pipe's pressure relative to atmospheric pressure). As a safety protection, everything is enclosed in another layer of air-proof material, and when a leak is detected the main air supply is shut off.

Special attention is made to make sure the average pressure in the pipe is the same as atmospheric pressure, since the piston motors depend on this to function.

(In real life, steam plants use direct current since there are a lot of losses due to condensation, and also since a lot of the point is transmitting thermal energy.)

Here is an entertaining take on some related questions, going into the history of how our modern ground systems evolved:

http://amasci.com/amateur/whygnd.html

I would be interested in your comments about that article, thanks!

Bill Beaty's site has a load of great articles on electricity (and much else), linked to here:

http://amasci.com/elect/elefaq.html

Also misconceptions about electricity taught in school:

http://amasci.com/miscon/elect.html

Or pages on generally believed science myths:

http://amasci.com/miscon/miscon.html

How transistors really work:

http://amasci.com/amateur/transis.html

How capacitors really work:

http://amasci.com/emotor/cap1.html

There's loads more great stuff. He's a pleasure to read - a guy who thinks for himself, and experiments - tries things out for himself.

So in the UK we have 3 wires, red, blue and green, the green is ground or earth, what purpose does this serve? Is it the same as ground as you call it in the US

I’m not an electrician but will take a stab at your answer. Please see if you can verify what I said so you don’t fry yourself.

If the green wire is connected between the receptacle and the outlet/switch then it’s “bonding” the two mental items. (Using the terminology from post). Most people refer to this as the ground wire or grounding the receptacle. The purpose, from my understanding, is to have a stable system. In the event the hot wire touches the receptacle.

If the red/blue/green are coming from the same cable (3 wires inside another sheath), then your green is the ground/neutral that connects to the electrical box itself, while the other two are hot/lead.

I posted a similar question above which may clarify things if the OP answers.

I'm from the US, so I'll add the caveat that this is answer is based on my foreign understanding and a British electrician may be better suited to answer this.

Your green wire would be what the US calls a ground wire, and what is more accurately called a Equipment Grounding Conductor (EGC). It's purpose is to bond all normally non-current carrying conductive parts of a system together to provide a ground-fault current path back to the source so that the overcurrent protective device can operate properly to open the circuit in the case of a ground fault. It is 100% a safety measure, and will operate as intended to open a breaker whether or not it is connected to the Earth.

Your red and black wires are both ungrounded conductors, and are the conductors which actually create the circuit that allows current to flow to power whatever you connect to the circuit. The circuit will work to provide current without the green wire. However, without the green wire, if either the red or black shorts to something, the breaker will not open and you have an enormous potential to harm someone or burn your house down.

> However, without the green wire, if either the red or black shorts to something, the breaker will not open and you have an enormous potential to harm someone or burn your house down.

This is not true. First of all only the red is hot, if the black shorts to something nothing will happen.

Second if there is a short strong enough to "burn your house down", the breaker most definitely will open.

After researching the UK electrical system, I realize I was mistaken on one regard. I thought they used 2-ungrounded conductors, but I was mistaken. The red wire is the only ungrounded. This would be the same as the black wire in a US house. The red wire is their grounded/neutral. This would be the same as a white wire in a US house.

I am correct on the EGC, though. It works the same as the EGC in the US. If a ground fault causes a house to burn down, then the breaker will absolutely NOT open. That's how the fire starts: current flows through something it shouldn't long enough to heat it up to the point that something ignites. If the breaker opens, the fire wouldn't have started in the first place.

The US NEC now requires arc fault circuit interrupters which are pretty darn expensive circuit breakers with built-in electronics to detect an arcing situation that wouldn't trigger a normal circuit breaker but can cause a fire.

I keep hearing electricians say that they can't detect bad splices which how most residential electrical fires start.

I've seen videos of these on YouTube and they seem useless!

Current British electrical standards require RCDs (GFCIs to you) to be fitted to new installations. So the RCD should trip 30 ms after the current starts flowing where it shouldn't. Plenty of installations won't be brought up to spec for decades of ever, however...

The reason for having ground is simply because if an electrician makes a mistake and reverses the hot and neutral, and you plug something in, the body of the device will now be connected to hot, and you'll get a shock.

By having a separate ground, it's harder to mess that up, and devices can connect the body of the device to that.

Inside the electrical box (mains) the neutral and ground are connected together, and are at the same potential.

The only reason they run separate wires is in case of of mistakes.

> The only reason they run separate wires is in case of of mistakes.

Erm no.

1. In case the neutral gets disconnected, so the frame isn't at the potential of the hot (no voltage drop across the device)

2. So the return voltage drop in the white wire isn't on the body of the device. ~20 volts between adjacent high-current appliances on different phases probably wouldn't be terribly dangerous to most people, but it is sloppy.

3. For GFCI/RCD, it separates bona fide return currents from "accidental" return currents. With only two wires, dropping a toaster into a (PVC-drain) bathtub wouldn't trip. (Erm, I just realized most toasters are actually only two wires. Well uh, don't do that).

If that "toaster circuit" is protected by a GFCI, it will indeed trip with only two wires. The GFCI is sensing the imbalance between hot and neutral (the rest went through the bath water). Both currents pass through the same toroid, but they are in opposite directions and normally exactly cancel, so no flux is induced in the toroid. If there is an imbalance, a current is induced in a secondary winding. This will trip the GFCI internal breaker. Note that this does not involve the green wire.

Except there will be no imbalance if there is no other path for the current to travel by...

The geometry of the electric field in water should probably be similar to that in air (mostly contained to within the case), just don't drop both a toaster and a waffle maker in at separate ends.

The resistance of slightly salty bathwater is much lower than air. While baths are not often plumbed with metal pipes these days, one can still expect some current flow to ground. It only takes milliamps of stray current to trip a GFCI.

Milliamps sounds like a tiny amount in the context of home power, but 5mA at 120V is still only 24kOhm. This is actually pretty low to just assume away. I'd agree there could be some water left over in the drain pipe from the last bath, splashes making a path to the spigot, or something like that. I just wouldn't count on it.

Iirc gfci's are nec code allowed on non grounded lines because they will still trip, but you have to test with the gfci test button because they're only tripping on current difference, not like a breaker.

Yeah, IIRC the "proper" way to upgrade an ungrounded circuit to have a 3-prong outlet (without running new cabling) is to use a GFCI and label it "NO EQUIPMENT GROUND". This is considered safe, but is still ultimately a hack.

It assumes that any dangerous shocking current will be taking a path that isn't just coming back on its own white wire, presumably from taking an unspecified path to ground. This seems a reasonable assumption for safety (and the NFPA surely has looked at the data), but it doesn't really inform the above isolated-bathtub situation.

There is also the matter of unanticipated leakage current between hot and the equipment enclosure. The EGC/green wire keeps the enclosure close to ground potential.

In the UK we used to have red, black and green, but now we have brown, blue and green/yellow (I think this is the same as rest of EU). I presume this is because of colour blindness reasons as there are folk who can't distinguish red and black that well.

Hopefully when Brexit is complete, you can return to the true colors as god and the Queen intended.

I don't have a specific question, but can you recommend a good, authoritative resource on learning to work with household electrical installations correctly, both theory and actual installation? Something like The Art of Electronics and Learning the Art of Electronics, but for general household electrical.

I do quite a bit of DIY electrical in my home, both because I enjoy tinkering and because I live in a developing country where most of the electricians you can hire is better described as "practiced amateur" rather than "professional," and I've gotten by with online resources (and the fact that my house is made from non-combustible materials) but as you point out in that post, there's a lot of conflicting or downright wrong information.

I don't want to become an actual electrician, but aside from just being interested in the topic, I'd at least want to know for sure that what I'm doing is safe and to be able to check the work of anyone I hire for obvious faults. I'm planning a solar installation in the near future, so I'd like to be well informed on how everything is supposed to work.

For a machine, is it generally a good idea to distribute a common power bus voltage like 48VDC to independent modules (embedded computers, servo controllers) for them to stepdown inside the application module (12vdc, 24vdc) or should the power be stepped down early at the distributor module? What are the advanced tradeoffs?

Generally the higher the voltage you distribute, the lower the current in the wires, allowing you to use thinner wire and reducing I^2R losses.

With this said, having multiple point of load converters may have impacts on efficiency and will add to the cost of the system.

There's one thing I don't understand between hot and neutral. How do the electric turbines upstream at the power plant only act on what becomes the hot wire? By my understanding of how motors work the magnet causes a positive current on the whole coiling/wire, meaning a positive current on one side is matched by an equally powerful negative current on the other. As both wires on the consumer end eventually join to the ends of that coiling, than wouldn't that mean both wires switch between hot and neutral everytime the current changes direction? Also, as all household wiring is in parallel off the mains, wouldn't that mean a closed circuit on one branch would turn the neutral wire on another branch hot as they effectively become one wire? How does the electric "push/pull" get separated onto only the hot line?

The electric generators aren't connected to your neutral wire. They deliver three phase power with three different "hot" lines, and then only one of those phases gets pulled down into your house electricity.

Yes, just as you describe, but the addition of a transformer allows you to reference the neutral to ground, therefore the hot swings above and below ground. It’s a little more complicated than what I described, as your breaker box is 240V split phase, but the transformer outside your house allows for a ground reference at its center tap.

If you have ever made the mistake of connecting a diode rectifier directly to AC mains, then grounding the negative output, you’ll blow a breaker. A 1:1 isolation transformer is still needed.

> than wouldn't that mean both wires switch between hot and neutral everytime the current changes direction?

The neutral line is connected to Earth at some point. The voltage difference is created by a transformer upstream which just pushes a 240V (or whatever) difference between live and neutral, and since neutral is tied to Earth, it stays at ~0V while the live line stays at ~240VAC.

You're right, though, that any current pulled from the live line goes back through the neutral line and they both end up with alternating current going through them. The voltage only alternates between live (or 'hot') and neutral relative to each other. This voltage difference between the two lines is generated by a transformer and there doesn't have to be any connection between live/neutral and anything else (at which point you can just think of them as live-1 and live-2) but the neutral line is connected to Earth at some point (although depending on load and the distance to the earthing point your neutral will actually have some AC ripple on it).

Here's a good one: A couple of years ago I borrowed a small camper trailer from my grandparents as I was going on a week camping trip. I towed the camper over to my house and left it hitched to the truck.

I then plugged the trailer into an outlet on the side of the house so I could run the air conditioner while I packed it. There was an issue with the brake lights not always working on trailer also. I crawled under the truck to trace the wires (they had been spliced under there before so I wanted to check that area first).

While on my back I rested my forearm against the frame of the truck and then felt a small "pinch". I thought a bug bit me at first, but seeing nothing there, I touched my forearm back again to the frame and felt it again. I realized it was electricity. I had a multi meter handy and I stuck one end of the probe in my mouth and touched the other end to the frame. It read 50 volts a/c. I went and unplugged the trailer from the outdoor outlet and ran it through a window and plugged it into an outlet inside and the issue went away.

So what's up with that?

Inproperly wired outlet? Hot swapped with neutral?

Or floating ground?

> So what's up with that?

Usually 50 VAC means the ground is open.

A thing about most ordinary transformers[1] is they wrap the primary and secondary together. Fast, cheap, and more efficient. The result is there is a large amount of capacitive coupling between the primary and secondary. Since one side of the secondary is grounded it's like connecting 1000-5000pf between hot and chassis ground. If chassis isn't grounded then it floats at 60VAC.

[1] For transformers with more isolation they wrap the primary and secondary on top of each other. Less capacitance that way. Medical grade transformers have physically separate windings.

Well, seems like you should do an AMA on Reddit!

This is a great introduction to protective earthing.

The only nit I'd pick is with the bit at the end about AC. The voltage on the neutral wire isn't due to inductive coupling, it's due to voltage drop along the neutral since neutral carries the same current as live. You'll always get this if anything's pulling significant current, earthing can't completely fix it, which is why you can't ever depend on neutral being the same voltage as earth, or as another neutral elsewhere (that's why we use dry contact relays to send signals between different pieces of kit - if they're getting neutral from different places then you can have a significant voltage difference between "neutral" in one cabinet and "neutral" in another, enough to blow up ELV gear!)

There's also the fact that, any time you use a transformer and don't earth both sides, you end up with a 'floating' AC circuit that will end up at some arbitrary voltage compared with earth. I've made this mistake before and been bitten by a stray (high impedance) 50VAC, not deadly but enough to get your attention. :P

How do you test that the "earth" connection is really good enough to redirect the current in that direction when need be?

I have been studying Grounding vs Bonding these past few weeks, great timing to see this HN discussion.

Some training videos for pro electricians and US National Electric Code compliance helped me.

https://youtu.be/qNZC782SzAQ

Basically, bond everything together if it can carry a charge. Ground that bond at only one place in an electrical service.

More than one "ground" sets up a situation where a potential difference can develop between the various "grounds". Ground Fault detection may no longer work in that case. Which is not good.

(The presentation is weirdly sexist in offhand comments, which I often encounter in trade work of this kind, but I do believe that's changing. Slowly. The USA needs more people who know how to build things.)

> The USA needs more people who know how to build things.

USA? The WORLD needs more people who know how to build things.

Watching this video (myth #3) brings up a bit of a semantic peeve. And that peeve begins with how the NEC narrowly defines the term "grounding" as only applying to the electrical system itself. Then electricians apply that narrow definition to every use of the term, acting like someone is wrong because they say they're "grounding" something by hooking up it up to an EGC.

But don't worry, they'll teach you this special term "bonding" that refers to the latter. Except it doesn't. I already used a synonym of "bonding" - "hooking up". You "bond" the black wire to the gold terminal on a receptacle by tightening the screw.

> Basically, bond everything together if it can carry a charge

It is perfectly fine to say that you are grounding those things together. Connecting something to an EGC is indeed "grounding it" - just not in NEC land which is focused on getting the EGC grounded. If you say "this washing machine needs to be better grounded", that doesn't mean it needs an immediate connection to earth, but is rather talking about its path to earth via the building's electrical system.

I thought multiple ground rods are code in many u.s. places.

Beware as this explanation does not apply to every country in the world, in France for residential systems at least, the protection (green/yellow) wire is never connected to the neutral (blue) wire, only to earth with a metal rod at the house. You would get a shock if it was somehow disconnected from the rod/earth. Neutral is also not bounded to earth inside the house. Be safe!

"Since the ground-fault current path has inherently much higher resistance than the neutral wire, the amount of current flowing through the circuit jumps enormously during a ground-fault."

This is incorrect (probably a typo) - the GFC ensures there is a very low resistance path to source, which ensures a current spike that is sufficient to blow the fuse.

Yeah, that was a typo on my part. I didn't notice it at first, and it's now been so long since I posted it that I would feel disingenuous editing it. I think someone pointed out the typo in the comments.

I encourage you to fix the typo; lots of HN exposure means lots of new readers without context to any time that has elapsed since your (really good) article was first posted.

It's very common to edit your typo but still leave it there either crossed out, and/or leave an EDIT: and explain what/why you changed it.

It's perfectly okay to ensure your text is communicating accurate information - in fact leaving it may undermine your explanation as certain folks will focus on the error (or others won't notice it and learn something incorrect).

The details of this also vary based on where you're at. In the US, neutral is bonded to earth at every house I think. In Europe, neutral is only bonded to earth at a local power station.

Yeah, this is 100% based on the US electrical system. I am not familiar enough with how other countries do it to speak intelligently on their systems.

The US system is somewhat unique and lacks heaps of safety features: I am guessing mostly because a 110V shock is far less dangerous than a 240V shock (Although higher amperages mean higher risk of fire?).

For example for UK, EU and NZ etc: sockets and pins have a variety of features to avoid touching the phase (live) wire with say a screwdriver or knife in hands of a child.

For example, all new circuits installed in New Zealand since 209 must have an Residual Current Detector (the actual rules lead to multiple RCDs per main power board). This will cut the circuit if someone does manage to touch a live wire somehow (current from phase to neutral doesn't match, because some of it is grounded through your body, and the mismatch triggers the breaker. Wayyyyyyy safer than a fuse!

In the US there is tamper proof outlets and GFCI is equivalent to RCD. Its pretty much required for new houses though older houses may vary on what they implement.

Are tamper proof outlets mandatory? The UK has had shutters on the phase that is opened by the earth pin forever. Which is why their plugs are caltrops, and have a plastic earth pin if double insulated (only phase and neutral wired, with no earth wire).

I don't think I have ever seen a US plug where the pins have a non-conductive protective covering at the base of the pin (I think all modern UK and NZ male plugs have that. Although I'll be honest that I think the NZ solution makes plugs less safe due to risk of pins bending then breaking and leaving a live metal pin exposed in the socket).

EU sockets are mostly recessed from what I have seen (some NZ sockets are, but the recess is still is often unusable with old plugs, so is not common yet, but will become so as more plugs are sold that fit properly).

> Are tamper proof outlets mandatory?

Its required by current US NEC code but older house are grandfathered in unless they do major renovation. Also some exceptions like outlets high enough to be out of reach of kids.

Its possible that the UK/NZ plugs have better designs. The code improves step by step here.

> The UK has had shutters on the phase that is opened by the earth pin forever. Which is why their plugs are caltrops

No, bad design is why they are caltrops. Over here in the Netherlands shutters are similarly required but none of our plugs need a third pin. The shutter is a see-saw construction that works in such a way that you need to insert both prongs at the same time for the shutter to rotate out of the way.

That said our grounded plugs are still caltrops for no real good reason, though less so than the UK ones. They tend to fall with the prongs to the side instead of prongs-up.

Not quite, there are different systems in service depending on region and installation era: https://diy.stackexchange.com/questions/134392/in-germany-sh...

This blew my mind. So the white wire (neutral) is actually what I thought the ground was (connected to earth) and the “ground” wire (usually bare or green) is a conductor for clearing ground faults and isn’t connected to earth.

Marvelous explanation.

Thanks! In my 11 year experience in the electrical industry I have come to realize that there is a LOT of misinformation about how electrical systems work, and even the most experienced veterans in the field often don't really understand it.

So what would be the potential consequence if I didn’t connect the “ground” wire (green or bare) to the little brass ground screw on (let’s say) a ceiling fan when installing it?

From my understanding, if the hot wire (black) were to come in contact with a piece of metal on the fan it would build up electrical charge and or allow for current to pass through the piece of the fan. Then if you were on a ladder fixing the fan and touched that piece of metal you would become a path for electricity to pass through to a lower potential.

That's exactly correct. You would get shocked if you touched the fan.

The way you bond is codified too. The source feed is wire-nutted with a pigtail connected to the box, and to the ground of the device in parallel. If the bonding was daisy chained it could be interrupted when servicing or changing.

If you do not properly bond the ceiling fan (that is, connect the "ground" wire) then the circuit breaker will not trip if there is a fault to ground on the fan.

That is, if the hot wire touches a metal part of the fan which is not supposed to be energized (say the metal housing), and the fan is not properly bonded, then there will be no ground-fault path back to the source to cause the breaker to trip. This could mean that the housing of your fan would be energized with (in a home in the US) 120V, and would shock anyone who touches it.

Technically, would it only shock you if your body provided a path to "ground" or lower potential? For example if you were wearing insulted boots and touched the energized piece of metal you would not feel a shock, correct?

Please don't insult your boots.

hahah, they piss me off sometimes though

Yes. That's why this is possible, here the insulation is the air between helicopter and ground: https://youtu.be/DPNK7bc2qvM?t=125

If that guy was connected to the earth he would get fried. Note that you still have to equalize whatever potential there is between you and the system, like they do in that video. Same reason why you should ground yourself and the hardware before working on your PC. Ever had a tiny spark hit you when you touched something (neither you or that "thing" being connected to the electrical grid, just static electricity e.g. from walking on the carpet)? But without connection to ground the only thing that happens is equalization of the potential between you and the target but no further flow. If you are connected to ground on one side and to a source that keeps creating an electrical potential, like the electrical grid with a power plant somewhere, there will be a continuing flow. If you are not connected to ground there won't be.

Correct. But if you touched the energized piece of metal with one hand and accidentally touched a grounded conductor with the other hand, you'd be dead. This is why in the rare circumstances when I have to work on an energized high-voltage circuit (don't do this!) I observe the "keep one hand in your pocket" rule.

So when installing any electrical fixtures at home I should always connect “ground” (bond the fixture) first. I’ve been connecting it last (always with the power off but accidents can happen). This could lead to an injury if someone flipped the breaker on and I accidentally touched hot to myself or the case because the breaker wouldn't trip.

Absolutely! Yes!

First, NEVER work anything hot. There is no such thing as a safe voltage to work energized. Even a simple home receptacle circuit can easily kill you (and it happens with alarming regularity).

That said, ALWAYS work something AS IF it were hot (for the exact reason you said, accidents happen). This means connecting the safety measures first. The safest order to connect wires is: ground, neutral, hot. The safest order to disconnect wires is the opposite: hot, neutral, ground.

> This could lead to an injury if someone flipped the breaker on and I accidentally touched hot to myself or the case because the breaker wouldn't trip.

Half right, half wrong. It would not short if it touched the metal box, and would not trip the breaker. That part is right.

However if it touched YOU not having the ground wire connected actually helps you. It means the electricity does not have as a good of a path back to the box, and you'll get a weaker shock.

With the ground wire you'll get a much higher shock, but it won't trip the breaker!! Humans have too high resistance to trip breakers, you'll just keep getting shocked.

So it's kind of a wash - with the ground, it's more likely to just short against the box and trip.

Best is just to always wire things as if they are live. Only touch the wire if you have to, always use insulated tools, even if the power is off.

In industrial and commercial settings, there are "Lockout/Tagout" rules. These prevent -- either through policy or physical means -- somebody from accidentally flipping a breaker back on when maintenance is occurring.

They are both connected to earth at the same point. The big difference is that "neutral" has current flowing through it during normal operation, but "ground" only has current flowing through it in a fault condition.

Right because the hot/black is connected to the neutral/white through the load completing a circuit so current flows. The “ground”/green is only connected to the chassis so no current flows in normal operation.

> conductor for clearing ground faults and isn’t connected to earth.

> Marvelous explanation.

It might be Marvelous, but it's not actually true. The ground wire IS connected to the earth!

I don't want to be rude, but I think the OP should take this post down, it has a LOT of mistakes.

TFA is a bit more abstract than you realize, and perhaps that abstraction could mislead the unwary. It is technically correct, but makes a distinction you've missed. If a gremlin cuts the (green) wire between my breaker box and my ground rod, the system is no longer "grounded to Earth", so it wouldn't handle lightning strikes well. However, the "EGC" in every house circuit is still bonded to neutral at the box, so breakers will still trip when the wire inside my clothes dryer rubs against the metal body. That is the function of the EGC. The EGC itself does nothing to protect against transient earth potentials like lightning strikes. That is the job of the grounding rod etc. Just because both wires are green, doesn't mean they're doing the same thing.

In the USA, there's no green wire between the service panel and the ground rod. The green wires end at the ground bus in the panel, which is bonded to the panel case. A heavy, bare copper wire leads from the ground bus to the ground rod.

The neutral bus is either bonded to the case with a green grounding screw or is integral with the ground bus.

See also "Why Three Prongs (1996)" discussed recently:

https://news.ycombinator.com/item?id=20378852

Also, be aware of grounding requirements when using a generator:

https://cpower.com/PDF/InfoSheets/44.pdf

Why are there three wires in new receptacles? From my understanding, new code has you have to have a 3+ground but why is that safer? Did they just add an extra wire so you can have a hot, lead, and neutral to ground all the receptacle? (Plus the green/bare to bond the receptacle to switch/outlet)

Also, for home owners, what’s the best resource to learn common sense basics of electrical work (besides reading the city/states codes). Are there classes for such cases?

The three wires in any typical household circuit are your Ungrounded Conductor (hot wire), Grounded Conductor (neutral wire), and Equipment Grounding Conductor (EGC/Ground Wire). The hot and neutral are what actually makes the circuit work. The EGC/ground is a safety measure that creates a ground fault current path which enables overcurrent protective devices (like breakers and fuses) to operate.

As far as learning common electrical, I'm sure local organizations in your area will have some kind of classes. Probably local hardware stores? I don't know. I learned through trade school and on-the-job training. As an electrician, I will self-servingly tell you that you should always hire a licensed electrician.

> The EGC/ground is a safety measure that creates a ground fault current path which enables overcurrent protective devices (like breakers and fuses) to operate.

This is NOT TRUE!! You've said it so many times in this thread, and it's just not true.

A lot of people are tying to correct your mistakes, but you're not fixing them. Please do so, you are misleading a lot of people.

Perhaps you could explain it correctly?

You can trip a breaker without the ground wire being involved. The ground wire does help with one failure case, but that's not the only way that your circuit breakers/fuses protect you.

Are you thinking about 240V circuits which used to use 3-wire circuits but now use 4-wire? The US run split-phase power to homes with +120V and -120V hot legs and neutral connected to center tap of transformer. Each 120V circuit uses one of the hot legs.

For 240V, can use 3-wire circuits by getting 240V between the +120V and -120V legs. There are 3-wire grounded plugs (hot-hot-ground) and non-grounded (hot-hot-neutral). Modern 240V wiring uses 4 wires, hot-hot-neutral-ground, which allows making 120V for low-power electronics from one hot to neutral.

It's not safer, it's more useful. You run out the hot, neutral, and switched hot. That lets you use a light switch to control an overhead light/fan without killing the fan.

One of the few times I wish I had a reddit account, just so I could give an old post some $REDDIT_KARMA_POINTS (whatever currency they use). That is an outstanding explanation that cleared up some things for me that I do because electrical code, not because "knows the 'why' of what he's doing".

Maybe it offers some solace to know that old posts (older than six months) cannot be voted on.

This section is plain wrong:

"The ONLY purpose for the EGC (or green wire) is to clear a ground-fault (clearing a ground-fault means tripping a breaker or blowing a fuse) in the 'oh shit moments'. It has absolutely NOTHING to do with the ground or the Earth and will work exactly as it is intended to regardless of whether it is connected to the Earth or not."

If the EGC is floating, no current will flow in a ground fault and the breaker never trips.

If the EGC is floating, it's not an EGC. EGC is tied back to Neutral at the distribution point. Its sole purpose is to carry fault current.

And the neutral is tied to (earth) ground. Otherwise all metal devices would shock you if you touched something earth referenced, such as a radiator, at the same time.

Edit to clarify: In principle, only the voltage difference between Live and Neutral matters for running a device. So theoretically one could construct a building supply where Neutral is not at earth potential. However, EGC has to be at earth potential, otherwise there would be a voltage between any metal electronic device and earthed objects, such as radiators or your feet standing in a puddle. This means that in practice both Neutral and EGC are always physically connected to the earth in some way and have to be for the system to work properly.

I'm not trying to be argumentative but you are misinformed. "Floating" means that the system is not connected to ground. That is, there is no system/main bonding jumper. There is no connection to a grounding electrode conductor. This has nothing to do with the Equipment Grounding Conductor, though. I don't know what you mean by a "floating EGC".

The entire purpose of the EGC is to bond all normally non-current carrying conductive parts together to provide a ground-fault current path back to the source of the circuit. If the EGC is not connected to some non-current carrying conductive part, then it is not installed correctly, is against code, and is a safety hazard. If the EGC is bonded to everything properly and there is a fault (say an ungrounded wire touches a metal box), then the current will flow through the ungrounded conductor, through the metal box, through the EGC (which is bonded to the box), back to the source (usually a transformer), across the windings, and eventually back to the breaker that controls the circuit. This will build up enough current (usually VERY fast, like milliseconds) and the breaker will open. The Earth has nothing to do with this.

He's right and you are not.

If you don't bond neutral and ground at the main, it's still bound at the transformer.

> then the current will flow through the ungrounded conductor,

No, it will flow through the grounded conductor.

> back to the source (usually a transformer), across the windings, and eventually back to the breaker that controls the circuit

The flow of power is the same via ground as it is via neutral.

> This will build up enough current (usually VERY fast, like milliseconds)

There is no such thing as "build up enough current", current does not "build up".

The milliseconds has to do with built in time delays at breakers, it's not a function of the electricity.

> The Earth has nothing to do with this

The actual Earth is used as a conductor, so yes, it definitely has something to do with this.

If you don't bond neutral and ground at the main, it's still bound at the transformer.

That would be a pretty good trick. The transformer is outside my house on a pole. Two wires run to it, the hot and the neutral. The "ground" doesn't leave the house. (Other wiring schemes exist, but this is standard USA residential.)

[EDIT: brainfart, see helpful correction below. still no "ground" at the pole...]

> Two wires run to it, the hot and the neutral

Standard US residential is 240V split phase, with two hot and one neutral conductor from the transformer. So three wires, minimum. Unless you have a very old feed.

Ah, yes, well spotted. That was an oversimplification. Two hots, one neutral, zero "grounds".

> still no "ground" at the pole...

Yes there is. The neutral is bonded to ground, via an actual metal pole in the ground.

In your house is the same thing: The neutral (and the ground) and attached to a metal pole in the ground.

And the earth itself completes the circuit.

FWIW "floating" has a more general use outside of the electrical trade. I read "if the EGC is floating" as "if the EGC is disconnected from where it normally is connected", which does make the original sentence true (just not interesting in the context of discussing why systems are generally grounded).

Also the EGC does have more purposes than simply clearing fault current - for example carrying away leakage current from the chassis of something with a switching power supply where the output can't be completely isolated to meet emissions. Ever been shocked by a laptop with only a two-prong AC adapter?

IMO this whole subject is a minefield of disagreements due to terminology, when really all questions are answered by drawing out the schematic of a typical electrical system and looking at the loops (circuits). For instance in the typical ground fault, the fault current returns to the distribution transformer in parallel through all of: your service's neutral, your grounding rod, your neighbors grounding rods/service neutrals, and your other leg of the split phase via turned-on devices. This seems like a lot of unrelated details to memorize until one draws it out.

> EGC (which is bonded to the box), back to the source (usually a transformer)

EGC is bonded to the panel enclosure at the ground bus. The neutral bus is also bonded to the enclosure with a ground screw. So, the EGC is directly connected to the neutral bus and a fault between hot and a bonded enclosure is routed via the EGC to neutral, tripping the breaker.

I'm confused by that as well. The green wire needs to touch the earth, whether by being bonded to neutral at the distribution point or simply by touching the earth somewhere, even through a concrete wall.

I've seen houses where the earth wiring is completely independent and terminated in a metal rod buried outside. Seemed to be up to local code, breakers worked as they should.

So I think you should define what you mean by "green wire". The NEC allows EGCs, GECs, MBJs, and SBJs to all use green wire. They all serve different purposes, and only the GEC is connected to the Earth. Whether or not the GEC is present the EGC and MBJ/SBJ is what allows breakers to work properly.

The MBJ/SBJ is what bonds the neutral. This doesn't have anything to do with the Earth. This bonding has to be done anywhere between the origination of the circuit and the first overcurrent protective device.

The GEC is what connects to the Earth. This is used to stabilize voltage and to eliminate a difference in potential between metal parts during a high frequency event (like a nearby lightning strike or a transformer blowing up). This has nothing to do with the operation of breakers.

Reply on news.ycombinator.com