I really don't see the use for the application, Because you have to deliberately put a load across line to earth to detect if the earth wire is connected to common which anything more than a simple neon is going to trip a GCFI circuit detecting current flow on earth.
Here is a simple schematic of an outlet tester to show how they test it.
View attachment 318485

Can you please tell how neon works in fault detection. I might not be aware of this but a little idea can be helpful.

 

I tried the way you said. Can you have a look at this.

I am not sure why you have two CTs in the circuit and if you are wanting to detect small currents (less than will trip and RCD / GFC) then 1000:1 CTs are going to make that a super small signal.
Other than that, the CTs (if I am understanding crutschow correctly) should be in the "earth" line not the active / hot line.
Also to dispel a myth or two, the reason the neutral and earth are connected together is to stop the active and neutral floating off whenever a storm could passes over and the reason this connection is made at a single point in an installation which is on the upstream side of circuit breakers / fuses is so that in the event of an active to earth hard fault, there is maximum fault current and the breaker / fuse will open quickly. If the neutral and earth wires are not connected together then the fault current has to travel through the dirt back to the origin (street or pole transformer) or wherever the earth neutral connection is made, then the fault current is less, the 'earthed' part which is presumably touchable by the user will be at a fairly high voltage for a while until the breaker trips or fuse melts. Not a good thing.
If you do insert a CT into the earth line then you should be aware of the fault current that your CT will have to carry. I can only speak for my local (Australian) domestic installations but I assume they are not so different elsewhere. The minimum fault current at a GPO (power outlet) is 3kA. And that is only if the kiosk (transformer) is as far away as they can be. If you are unlucky enough to have the kiosk right at your door, then the fault current is 10kA. These are hard active to neutral or active to earth faults.
You might wanna put some surge protection of the output of the CT and on the input to the CT as well because no matter what you do, the CT is unlikely to handle the fault current gracefully.
If I were you and assuming I understand what you are trying to do, if you install a medical grade GFC/RCD which trips at 10mA instead of the usual 30mA then that will give you a binary indication of the earth current without making anything unsafe or non compliant with the wiring rules.

 

How can the OK light signify a fault?

Yeah, I don't think that little circuit of 3 neons works quite the way it is supposed to.

 

I am not sure why you have two CTs in the circuit and if you are wanting to detect small currents (less than will trip and RCD / GFC) then 1000:1 CTs are going to make that a super small signal.
Other than that, the CTs (if I am understanding crutschow correctly) should be in the "earth" line not the active / hot line.
Also to dispel a myth or two, the reason the neutral and earth are connected together is to stop the active and neutral floating off whenever a storm could passes over and the reason this connection is made at a single point in an installation which is on the upstream side of circuit breakers / fuses is so that in the event of an active to earth hard fault, there is maximum fault current and the breaker / fuse will open quickly. If the neutral and earth wires are not connected together then the fault current has to travel through the dirt back to the origin (street or pole transformer) or wherever the earth neutral connection is made, then the fault current is less, the 'earthed' part which is presumably touchable by the user will be at a fairly high voltage for a while until the breaker trips or fuse melts. Not a good thing.
If you do insert a CT into the earth line then you should be aware of the fault current that your CT will have to carry. I can only speak for my local (Australian) domestic installations but I assume they are not so different elsewhere. The minimum fault current at a GPO (power outlet) is 3kA. And that is only if the kiosk (transformer) is as far away as they can be. If you are unlucky enough to have the kiosk right at your door, then the fault current is 10kA. These are hard active to neutral or active to earth faults.
You might wanna put some surge protection of the output of the CT and on the input to the CT as well because no matter what you do, the CT is unlikely to handle the fault current gracefully.
If I were you and assuming I understand what you are trying to do, if you install a medical grade GFC/RCD which trips at 10mA instead of the usual 30mA then that will give you a binary indication of the earth current without making anything unsafe or non compliant with the wiring rules.

D'Oh! The 1000:1 CT will make the signal bigger! What was I thinking!? It has been a long week already and it is only Wednesday. Or maybe a senior moment. Meh.

 

Shouldn't Neutral and Earth Ground be bonded at the service entrance? Here in the US that is how the NEC has things wired.



View attachment 318497

AC receptical testers are actually pretty inexpensive, like these just as a single example.

Ron

Here's an example from my experience.

Our prior house had a subpanel and when the pre-purchase inspection was done, the inspector noted that the neutral was tied to the earth ground in BOTH panels.

The fact that it shouldn't have been didn't change the fact that it was.

Measuring the current in the earth ground on the 'house' side of the ground-to-neutral bond in the main panel "probably" would have detected this, since it is almost certain that "some" of the current from the subpanel would travel back to the main panel via the ground conductor between them. Not guaranteed, but probable. However, in our case, the ground wire at the subpanel came from the main panel (which is probably how any such incorrect arrangement would have been wired). But if the subpanel had been bonded to a ground that was separately grounded to earth, there would be no current in the ground wire in the main panel and you would have to look for a mismatch between total line current and total neutral current.

 

Can you please tell how neon works in fault detection. I might not be aware of this but a little idea can be helpful.

The outlet test light will only show how an outlet is wired.
Fault detection is monitored by the GCFI circuit breaker.
I guess since they don't use the green wire, you could put your current transformer there, But you you would have to put it somewhere other than inside the breaker panel or the outlet box (a junction box inline)
But I really don't see the use for it even if you don't have GCFI circuit breakers.

 

Here's an example from my experience.

Our prior house had a subpanel and when the pre-purchase inspection was done, the inspector noted that the neutral was tied to the earth ground in BOTH panels.

The fact that it shouldn't have been didn't change the fact that it was.

Measuring the current in the earth ground on the 'house' side of the ground-to-neutral bond in the main panel "probably" would have detected this, since it is almost certain that "some" of the current from the subpanel would travel back to the main panel via the ground conductor between them. Not guaranteed, but probable. However, in our case, the ground wire at the subpanel came from the main panel (which is probably how any such incorrect arrangement would have been wired). But if the subpanel had been bonded to a ground that was separately grounded to earth, there would be no current in the ground wire in the main panel and you would have to look for a mismatch between total line current and total neutral current.

The NEC used to have a few things wrong with subpanel installations and this was one of them. The only bonding that should happen in a subpanel is Earth ground bonding of the green wire to a grounding rod and the white wires bus spliced and carry over to the main panel without being bonded to that earth ground. The only thing that hasn't been thoroughly worked out is how isolated ground subpanels should be installed. Personally, I think that an isolation transformer should be deployed so that balanced power plus isolation is achieved at those panels. But there isn't much demand for that in new industrial installations.

That subpanel might have been OK when it was wired that way, but now the NEC learned that they need to install them like the industrial electricians classically did and adopted that method across all installations.

 

The NEC used to have a few things wrong with subpanel installations and this was one of them. The only bonding that should happen in a subpanel is Earth ground bonding of the green wire to a grounding rod and the white wires bus spliced and carry over to the main panel without being bonded to that earth ground. The only thing that hasn't been thoroughly worked out is how isolated ground subpanels should be installed. Personally, I think that an isolation transformer should be deployed so that balanced power plus isolation is achieved at those panels. But there isn't much demand for that in new industrial installations.

That subpanel might have been OK when it was wired that way, but now the NEC learned that they need to install them like the industrial electricians classically did and adopted that method across all installations.

The problem isn't the NEC, it's that the subpanel was not wired to NEC code when it was installed (which happened in the early 2000s). There's nothing gained by blaming the NEC for some contractor's mistake.

 

The problem isn't the NEC, it's that the subpanel was not wired to NEC code when it was installed (which happened in the early 2000s). There's nothing gained by blaming the NEC for some contractor's mistake.

Don't know about that . But they gotten things wrong before in the past. And just like every other thing, there is politics involved.
So I would have to say, they didn't have that rule, and changed it.
Not like everywhere they use the NEC either. Rural areas are really bad that way.

 

Don't know about that . But they gotten things wrong before in the past. And just like every other thing, there is politics involved.
So I would have to say, they didn't have that rule, and changed it.
Not like everywhere they use the NEC either. Rural areas are really bad that way.

The rule regarding having a single point connection between the neutral and ground has been in place for decades -- I remember it being a point of emphasis when I took an electronics course in high school and we had to wire a mini-house which included a subpanel. The sad reality is that many people that work at that level have little actual understanding of what they are doing -- they follow a recipe they have been given without can concern for why those recipes are the way they are. Most of them never work with a subpanel, and so the recipe for all of the panels they install call for bonding the neutral to the ground. Then, when they do install a subpanel, they just simply follow the same recipe because they don't understand the reasons for the rules and, to them, a panel is a panel. They almost certainly were told about the different setup in a subpanel, though the reason may or may not have been explained at the time, but they have long-since forget that and so they just follow the recipe they think they have memorized.

 

As to the NEC I believe every 3 years they get together have a few beers and make changes.

Seriously though, every 3 years a new book comes out and changes are published. A recent change was running the neutral through work boxes allowing the use of wireless switches and similar. Ground bonding has been around a long, long time at service entry. Nothing new.

Ron

 

The NEC used to have a few things wrong with subpanel installations and this was one of them. The only bonding that should happen in a subpanel is Earth ground bonding of the green wire to a grounding rod and the white wires bus spliced and carry over to the main panel without being bonded to that earth ground. The only thing that hasn't been thoroughly worked out is how isolated ground subpanels should be installed. Personally, I think that an isolation transformer should be deployed so that balanced power plus isolation is achieved at those panels. But there isn't much demand for that in new industrial installations.

That subpanel might have been OK when it was wired that way, but now the NEC learned that they need to install them like the industrial electricians classically did and adopted that method across all installations.

If I understand what you are saying, it makes sense until you remember all the conductive pipes that are buried in the ground around suburbs and routed into every dwelling in every street. In Aus' a domestic installation has only one bonding point which is the main panel (there are only rarely ever sub panels in domestic installations). But the bonding is more than just earth to neutral and earth rod with a link between the earth bus bar and the neutral bus bar called the MEN link (Multiple Earth Neutral). We also have 'equipotential bonding' which is linking the earth bar to the water and gas pipes if they are conductive (copper) pipes.

This doesn't stop currents from flowing in the water mains in the pipe in the street (a thing that killed guys working on those pipes until they forced them to put a heavy wire between pipe sections before they disconnected the section of pipe). This does limit the voltage though if the connecting wire between the sections of pipe does not quite make contact.

Earthing, grounding and bonding gets really complicated the more you think about it, the more of the distribution network you take into account and what happens in a fault condition. I guess that is why there is no definitive answer for which way is best.

 

If I understand what you are saying, it makes sense until you remember all the conductive pipes that are buried in the ground around suburbs and routed into every dwelling in every street. In Aus' a domestic installation has only one bonding point which is the main panel (there are only rarely ever sub panels in domestic installations). But the bonding is more than just earth to neutral and earth rod with a link between the earth bus bar and the neutral bus bar called the MEN link (Multiple Earth Neutral). We also have 'equipotential bonding' which is linking the earth bar to the water and gas pipes if they are conductive (copper) pipes.

now these days a modern building don't have metal water pipes. I'm sure certain conditions has to be present or test on it to rely on a pipe ground by itself. But the grounds I see are the ground rod and the ufer grounding.

 

now these days a modern building don't have metal water pipes. I'm sure certain conditions has to be present or test on it to rely on a pipe ground by itself. But the grounds I see are the ground rod and the ufer grounding.

I never said the earth ground was made by connection to pipes. That is as I said an equipotential connection. the earth connection, I said, was a rod.
Yes, modern building practice tends not to use metal pipes for water or gas so these equipotential connections are becoming less common.

 

OK, if there is a valid need to detect that at some internal point the white neutral has contacted the green-wire safety ground, here is how to do it. BUT the green wire ground must not be grounded else-where. The scheme will be to have a sensitive current transformer, such as found in GFCI devices, watching the ground wire for any current. THAT is the simple method. The other, more complex method is to pass both of the mains sides, power and neutral, thru a similar sensitive current transformer, AFTER the point at which the neutral is tied to the safety ground circuit. THEN any leakage from neutral to ground will trip the detector. The only flaw here is that also, any leakage from the mains power to "ground" will also trip the detector. So there are the two schemes that will work, although both will also be triggered by leakage from the "power" side to "green-wire" ground.

And my professional opinion is that this "neutral to green" detection scheme provides no additional safety value at all.

 

When I left the industry in the UK, back then the domestic residence neutral could not come in contact with earth GND anywhere in the installation, it relied on the GND where it was tied to the distribution transformer neutral. And earth GND had to be resistance tested back to the transformer.
Then with the advent of mandatory RCD devices it was grounded at the residence panel.
In N.A and elsewhere, Industrial installations where a 120v/240v control is derived from a separate control transformer, it is allowed to locally earth one side of a secondary TXFR output point.

 

When I left the industry in the UK, back then the domestic residence neutral could not come in contact with earth GND anywhere in the installation, it relied on the GND where it was tied to the distribution transformer neutral. And earth GND had to be resistance tested back to the transformer.
Then with the advent of mandatory RCD devices it was grounded at the residence panel.
In N.A and elsewhere, Industrial installations where a 120v/240v control is derived from a separate control transformer, it is allowed to locally earth one side of a secondary TXFR output point.

That's the key, a separately derived system, often using a transformer. I have separately derived power in the house for my media room to eliminate ground and line noise generated at the main panel G/N bonding from AC appliances. I have a secondary grounding rod and ground for the separate secondary to the sub-panel with the secondary rod bonded outside the house to the primary rod(s) before the main panel connections.
https://forum.allaboutcircuits.com/threads/using-the-ground-as-a-conductor.149231/post-1275191

 

If the circuit is actually isolated, as shown, then what advantage is provided by connecting one side to any common point shared with the non-isolated system??
The explanation that I have received before is "that everybody does it", and that is not very effective in describing any benefit for either safety or noise reduction.
An adequately isolated system, with a correctly "grounded" electrostatic shield between transformer primary and secondary, should not have any common-mode noise or disturbances coupled in from outside.

 

And my professional opinion is that this "neutral to green" detection scheme provides no additional safety value at all.

Its why a GCFI doesn't use it and detects if all the current going out of the voltage leg returns on the white wire.

 

If the circuit is actually isolated, as shown, then what advantage is provided by connecting one side to any common point shared with the non-isolated system??

On a distribution transformer going into a building they earth ground the center tap to insure balanced power and establish a voltage reference from common to earth.