GFCI Neutral to Ground fault

[ripp1202]

How does a GFCI work:

GFCIs typically test for the following condition:

* A Hot to Ground (safety/earth) fault. Current flows from the Hot wire to Ground bypassing the Neutral. This is the test that is most critical for safety.
* A Grounded Neutral fault. Due to miswiring or a short circuit, the N and G wires are connected by a low resistance path downstream of the GFCI. In this case, the GFCI will trip as soon as power is applied even if nothing is connected to its protected (load) circuit.

To detect a Hot to Ground fault, both current carrying wires pass through the core of a sense coil (transformer). When the currents are equal and opposite, there is no output from its multiturn sense voltage winding. When an imbalance occurs, an output signal is produced. When this exceeds a threshold, a circuit breaker inside the GFCI is tripped.

To detect a Neutral to Ground fault there is a second transformer (left toroid in the illustration below) placed upstream of the H-G sense transformer (in the illustration above). A small drive signal is injected via the 200 T winding which induces equal voltages on the H and N wires passing through its core.

Neutral to Ground Fault Detection

* If N and G are separate downstream (as they should be), no current will be flow in either wire and the GFCI will not trip. (No current will flow in the H wire as a result of this stimulus because the voltage induced on both H and N is equal and cancels.)
* If there is a N-G short downstream, a current will flow through the N wire, to the G wire via the short, and back to the N wire via the normal connection at the service panel. Since there will be NO similar current in the H wire, this represents a current unbalance and will trip the GFCI in the same manner as the usual H-G short.
* If there is a H-H

[Incidently, a type A GFCI will detect a "hotted hot" <G> as well as a gounded neutral. If there is a parallel path path from the load side hot back to line side hot, it will trip via the same mechanism as the load grounded neutral trip. So, a GFCI won't work on a "double ended" circuit.

It works pretty simply when you study that circuit you pointed us to to. The second coil has as it's primary the ufiltered output of the full-wave rectifier. If a closed loop condition exists between any of the two wires going through the coil, this will induce a ~120Hz current in that closed loop. Ingenious!...

[Bill]

I had never heard of, or seen, or read about, a double-toroid GFCI. In fact, I took one apart the other day. It was a decent US-made unit (Leviton, I think), only about two years old, and it had only a single toroid.

I'm a little confused about the added value of the second toroid. Yes, it makes it possible to detect a neutral-to-ground short - but as we have continually seen on this board, even a single-toroid unit finds this problem. In either case, the result is a mysterious "nuisance trip" that is hard to diagnose. Does the double-toroid unit have a separate indicator light for a neutral-to-ground fault? That would certainly help.

The one difference I can see is that the double-toroid unit provides self-stimulus for the detection process, while the single toroid unit that I described relies on an external source of fault current - the TV set in my example.

I'm glad to see that I am not the only one who has noticed this problem, though, and that the manufacturers have already taken steps to deal with it. What was the source of the info you posted, and the diagram?

Bill

[ripp1202]

Some common neutral/ground problems/solutions:

1. An appliance. That could be the power converter, inverter, water heater, refrigerator, air conditioner, microwave, TV, etc. - anything that runs on 120 volts. Unplug all that have plugs, and try the GFCI. If it doesn't trip, starting plugging them back in one at a time until the problem occurs.

2. Neutral terminal shorted to the uninsulated ground wire in an outlet or at a switch. The ground wire is not insulated, and when the wires are wadded up and pushed into the outlet box, the bare ground wire can touch the neutral wire/connection point. The task is to remove each one and visually inspect it.

3. Neutral shorted to ground. The MH frame, aluminum skin, etc. are grounded. If a screw or staple penetrates the electical wire and touches any grounded metal, that will cause the GFCI to trip. The same if a sharp metal edge wears through the insulation. To find the problem cable, disconnect each of the neutrals from the MH electrical panel and check the resistance between each and ground with an ohmmeter. They should all read open, but there will probably be one that does not; follow that cable to the problem.

[ripp1202]

http://www.electronics-manufacturers...pter-gfci.html http://www.rhtubs.com/GFCI/GFCI.htm