GFCI's

Status
Not open for further replies.
B

byenzer

Member
Location
95621
Occupation
Self
On any GFCIs: WITH no load attached, if you touch the neutral pin to the ground pin, it trips the GFCI. WHY? The ground and the neutral are both at the same potential, so WHY?
 
byenzer said:
On any GFCIs: WITH no load attached, if you touch the neutral pin to the ground pin, it trips the GFCI. WHY? The ground and the neutral are both at the same potential, so WHY?
Click to expand...
Welcome to the forum.

Because GFCIs intentionally senses it, to make sure it correctly detects when the line currents mismatch.
 
There is a small load inside the GFCI from the sensing electronics and the little LED light showing either on or tripped.
 
byenzer said:
On any GFCIs: WITH no load attached, if you touch the neutral pin to the ground pin, it trips the GFCI. WHY? The ground and the neutral are both at the same potential, so WHY?
Click to expand...

It trips because there's extra circuitry within GFCIs to detect a grounded neutral condition as required by UL943.

The reason for this requirement is that a short (or a relatively low resistance) between the neutral and equipment ground conductor (EGC) on the load side of a GFCI will reduce the sensitivity to a ground fault. This is because such a connection will divert a portion of any leakage current going from the hot conductor to the EGC back through the load neutral conductor, instead of having all of the leakage current return through the ECG. Therefore within the toroidal CT of the GFCI this diverted part of the leakage current back through the neutral will effectively subtract from the actual leakage current, in terms of what the CT will actually be sensing. As a result a higher leakage current would be needed to trip the GFCI. To avoid this situation extra circuitry is present that will trip the GFCI if there's a N-EGC connection.

I'm aware of a few ways this functionality is achieved. I believe a common method is to use a second toroid with an oscillator circuit to apply a relatively low level, higher frequency common-mode voltage at the load terminals. As a result, grounding the output neutral terminal will cause a higher frequency current to flow through the neutral, which is then sensed by the first toroid and trips the GFCI.
Another method I saw described detected the extra loading that a N-EGC connection applies to a tuned circuit formed with a second toroid, and which reduces the sharpness of its resonance or "Q" factor. In any case UL943 does not specify how grounded neutral detection is implemented, only the perfomance specs it has to meet.

Above Larry said essentially the same thing, but with more concise wording. ;)
 
Last edited:
A signal is injected into the protected side conductors, if it is unbalanced (current path would be from protected neutral to EGC, back to main bond jumper, through the source and back on the unprotected "hot" it will trip same device at same trip threshold that trips for the supply voltage faults.

Neutral to ground fault protection has been part of listing requirements for a long time AFAIK.

May need to play with this some time. In theory if connected to an ungrounded source it should not trip - until there is a ground reference somewhere.

Of course any direct line to load side current will trip it - that is exactly what pressing the test button does.
 
byenzer said:
On any GFCIs: WITH no load attached, if you touch the neutral pin to the ground pin, it trips the GFCI. WHY? The ground and the neutral are both at the same potential, so WHY?
Click to expand...
Try touching it the other way, touch ground to neutral. ;)

It's likely ground loop detection. EGC's do not complete a ckt, unless in a fault.
 
byenzer said:
There is ZERO CURRENT FLOW.
Click to expand...

The basic GFCI function does not see current in the neutral to ground fault condition. Thus the basic GFCI function does not trip in this case.

As others have described, detecting a grounded neutral condition is a separate requirement detected in a different fashion.

Jon
 
winnie said:
The basic GFCI function does not see current in the neutral to ground fault condition. Thus the basic GFCI function does not trip in this case.

As others have described, detecting a grounded neutral condition is a separate requirement detected in a different fashion.

Jon
Click to expand...
kind of is detected in same fashion as other detected faults. No the basic GFCI function alone will not detect neutral to ground fault unless there is enough current flowing to ground to make it trip. (4-6 mA)

What is different is there is a signal derived from within the device and an equal voltage is injected on both hot and neutral conductors. As long as there is no faults to anything outside the protected conductors downstream of the device there is no current associated with this voltage. A neutral to ground fault will allow current to flow from that injection point, through the load neutral, through the fault point, back to the system/main bonding jumper and back to the supply side neutral of the device, creating an imbalance current of high enough level to make the main trip circuitry detect it and trip.

This system also would trip if you had load side hot fault to line side hot (same phase) in similar manner. Fault two phases you may or may not trip the GFCI but you are likely to also trip the OCPD.
 
FWIW:
Fluke 87 and T+Pro. EG is conduit that is only associated with this circuit. The T+ is not a low impedance tester but is lower than the 87.
No load on GFCI other than its internal Power LED. Leviton device.
GFCI On: 4.6-5mV N-EG with 87. Paralleled test connection of T+ does not change the voltage.
GFCI Off: 364mV N-EG with 87. Paralleled T+ drops voltage to 71.2 mV.
Set Min/Max of 87 to mA mode. AC. N-EG catches 13.68mA as it trips the GFCI
(Yes, I risked the $13 fuse just for this thread)

So there is current flow.
You're welcome.
 
ptonsparky said:
FWIW:
Fluke 87 and T+Pro. EG is conduit that is only associated with this circuit. The T+ is not a low impedance tester but is lower than the 87.
No load on GFCI other than its internal Power LED. Leviton device.
GFCI On: 4.6-5mV N-EG with 87. Paralleled test connection of T+ does not change the voltage.
GFCI Off: 364mV N-EG with 87. Paralleled T+ drops voltage to 71.2 mV.
Set Min/Max of 87 to mA mode. AC. N-EG catches 13.68mA as it trips the GFCI
(Yes, I risked the $13 fuse just for this thread)

So there is current flow.
You're welcome.
Click to expand...
This also helps prove that GFCI's won't protect you from shock or that they will limit current to 4-6 mA. They just will respond quickly to a fault of 4-6 mA or more.
 
kwired said:
This also helps prove that GFCI's won't protect you from shock or that they will limit current to 4-6 mA. They just will respond quickly to a fault of 4-6 mA or more.
Click to expand...
If you think about it, a shock is required for it to react.
 
LarryFine said:
If you think about it, a shock is required for it to react.
Click to expand...
Yes and no. Many (probably most) times a fault has occurred and tripped the device before current traveled through a person.

But if it is a person's making the necessary contact to cause the fault current then yes.
 
What if my hair has a bunch of static charge and I touch the N wire on load side of the GFI?
 
The capacitance of a person relative to 'Earth' is pretty small, so the charge associated with even a high voltage is pretty small.

I'd expect a discharge period of perhaps 1 microsecond (the standard human body model is 100pF and 1500 ohms, so a time constant of 1/6 us) but with a current that might peak at a couple of amps.

I don't know what a GFCI would do with that. My guess is that it would get ignored because of filtering.

Jon
 
FionaZuppa said:
What if my hair has a bunch of static charge and I touch the N wire on load side of the GFI?
Click to expand...
going to depend on how much current it can produce and whether it flows in a way to unbalance the sense coil of the GFCI.

If it does produce enough to trip, you are definitely going to feel it.
 
Status
Not open for further replies.
Top