Gfci wont trip with tester

Status
Not open for further replies.
I just tried a brand new out of the box P&S GFCI and it will not trip with a load side grounded to grounding connection. This device says it is in compliance with the new UL standard.
Don
 
don_resqcapt19 said:
I just tried a brand new out of the box P&S GFCI and it will not trip with a load side grounded to grounding connection. This device says it is in compliance with the new UL standard.
Don
Click to expand...
Interesting.

Not to ask the obvious or to be insulting ... but did your test setup include a path from the recepticle ground terminal back to to the unprotected "neutral line" side of the GFCI?

(i.e. the same path which would be provided via an installed EGC and main bonding jumper)
 
Not to ask the obvious or to be insulting ... but did your test setup include a path from the recepticle ground terminal back to to the unprotected "neutral line" side of the GFCI?
Click to expand...
Yes the EGC was connected and the device would trip when I put my wiggy between the protected hot and the EGC. It would not trip between the grounded conductor and the EGC. My wiggy is also continuity tester and it showed continuity between the grounded and grounding conductor.


Don
 
don_resqcapt19 said:
Yes the EGC was connected ... it showed continuity between the grounded and grounding conductor.
Click to expand...
Thanks for the reply.

I had success w/ U.L. today! :smile:

A Neutral-to-Ground fault should cause a properly-functioning GFCI to trip per the U.L. Std 943 (1 Feb 2006), Para 6.7.4 -- Grounded Neutral Test.

 
Last edited:
don_resqcapt19 said:
Yes the EGC was connected and the device would trip when I put my wiggy between the protected hot and the EGC. It would not trip between the grounded conductor and the EGC.
Click to expand...
Don, try a solid wire instead of the wiggy.
 
tallgirl said:
All the GFCI's I've ever tried to trip with a neutral to ground fault have done so.
Click to expand...

Yes if you have a 'bolted' connection between neutral and ground once you apply a load the GFCI will trip due to current imbalance.

Like Don I have never tripped one using a wiggy between neutral and ground.

Also any GFCI I have destroyed to look at only had one sensing coil but none of those have been very new.
 
The neutral-ground connection will trip GFCIs I have had before these newest ones (that trip or won't reset for line/load and other wiring errors). My point is it isn't that recent (like 1996).

I can understand a wiggy not tripping it, but it doesn't need a load, just a low enough resistance (I always used a short between neutral and ground to try it).

My understanding on why it does that: if you have some kind of connection between neutral and ground, that will lower the neutral current (split between neutral and EGC), and if a true ground fault lowered the hot current, the regular GFCI part wouldn't trip if this hot and neutral current were similar, even though there is a ground fault.
 
Jim,
A Neutral-to-Ground fault should cause a properly-functioning GFCI to trip per the U.L. Std 943 (1 Feb 2006), Para 6.7.4 -- Grounded Neutral Test.
Click to expand...
But only if the neutral to ground fault results in current flow, just like any other ground fault. Don
 
don_resqcapt19 said:
Jim,
[/I]But only if the neutral to ground fault results in current flow, just like any other ground fault. Don
Click to expand...

If the neutral is bonded at a point other than where the circuit originates there should be current flow between neutral and ground, even with no load on the circuit -- Kirchoff's Law.

Don't actually try this at home -- get a DVM and measure the voltage between neutral and ground. If it's non-zero, there will be current flow on the neutral. If you want to have some serious fun, find a receptacle at the end of a run. Load up that circuit. Get an old 3 wire cord, cut it about 1' from the plug, strip back the covering and connect the neutral wire to the ground wire. Plug that into the receptacle. Then put a clamp on meter around that wire. The amount of current which flows is left to the reader as an exercise.
 
Julie,
If the neutral is bonded at a point other than where the circuit originates there should be current flow between neutral and ground, even with no load on the circuit --
Click to expand...
Where does the voltage come from to drive this current? The grounded and grounding conductors are bonded together at the main bonding jumper. If you bond them at another location the only voltage that could be available to drive the current would be the voltage drop on the grounded conductor between the 2 bonded points. If this voltage is enough to drive ~5 mA then the GFCI will trip even where there is a single current sensor. If there is no load on the grounded conductor to the GFCI, then there is no voltage drop and there will be no current flow.
Don
 
don_resqcapt19 said:
Julie,

Where does the voltage come from to drive this current? The grounded and grounding conductors are bonded together at the main bonding jumper. If you bond them at another location the only voltage that could be available to drive the current would be the voltage drop on the grounded conductor between the 2 bonded points. If this voltage is enough to drive ~5 mA then the GFCI will trip even where there is a single current sensor. If there is no load on the grounded conductor to the GFCI, then there is no voltage drop and there will be no current flow.
Don
Click to expand...

1). Voltage Drop
2). Kirchoff's Law
3). Parallel Conductors
4). Subpanels

That's my answer and I'm sticking to it :)

The GFCI's neutral-ground fault detection works a different way, however -- it inserts a signal onto the grounded conductor and looks for the signal on the EGC.
 
After going set-by-step thru the U.L. test proceedure, I believe I have the answer (and we're all partially right).

Summary:

1) GFCIs MUST trip if threre is a neutral-to-ground fault.

2) The fault may be (but is not required to be) detected -- and the GFCI subsequently tripped -- w/o any current being tapped from the protected load-side hot/line terminal. (This is specifically the configuration in step 6.7.4.3(f) of the U.L. test proceedure)

3) If the device does not trip w/ a neutral-to-ground fault alone in the step above, it MUST trip when the test-procedure-determined "hot/line" current is applied)

*****

This test proceedure would give manufacturers some flexability in design, and would allow both of the sorts of behaviors we're observing.

My Cooper devices (and apparently Julia's) trip on simple detection of the neutral-to-ground fault alone, w/o any current from the load's "hot/line" terminal being required ... presumably by a detection device similar to that described in the Fairfchild document which somehow senses when a "loop" circuit is completed from the GFCI protected neutral, thru the EGC and Main Bonding Jumper back to the unprotected neutral-side of the device.

Don's P&S devices apparently don't trip on the neutral-to-ground fault alone -- presumably because they use a different type of detector -- but as long as they do trip when the test-specified amount of "hot/line" current is applied they still comply w/ the U.L. standard.

If anyone is interested in the reviewing the actual U.L neutral-to-ground fault test proceedure so that they might make their own interpretation, please I.M. me.
 
Julie,
I don't understand where the voltage comes from. The grounded and grounding conductors are connected togther in the serivce panel. They are at zero potential to each other at that point. If you extend both of them away from that panel without a load on either of them there is still zero potential between them at the end. Do you happen to live in a place where the GFCIs are fed from subpanel and not the main panel? If so then there is a voltage betweeen the grounded and grounding conductor at the end of the branch circuit even without a load on that branch circuit. This voltage will be equal to the voltage drop in the feeder grounded conductor.
The GFCI's neutral-ground fault detection works a different way, however -- it inserts a signal onto the grounded conductor and looks for the signal on the EGC.
Click to expand...
While some may work that way, there is no requirement that they use that method. My understanding of the UL standard is that the GFCI is only required to trip with a neutral ground fault when there is ~5mA of current flowing on that ground fault. A standard single sensor GFCI can do that, because the current at the point of the sensor is not equal in both the ungrounded and grounded conductor.
Don
 
All the GFCI's I have tried were P&S, and they all trip by connecting a wire from the neutral to ground of the receptacle. With no load I would expect no current to flow, but the GFCI still trips. I still think Don's case is using a Wiggy instead of a wire makes a difference.
 
Paul,
With no load I would expect no current to flow, but the GFCI still trips. I still think Don's case is using a Wiggy instead of a wire makes a difference.
Click to expand...
That could be, but I think the bigger issue is voltage drop on the grounded conductor between the main bonding jumper and the neutral to ground fault. If there is only a very small voltage drop, then the wiggy could limit the current below the 5 mA required to trip the GFCI.
Don
 
I tied it again with a new P&S GFCI. The power source is a sub-panel and I have 0.81 volts from the grounded conductor to the grounding conductor. Neither my wiggy nor a soild jumper will trip the GFCI with a load side grounded conductor to grounding conductor fault. A solid wire fault from line side grounded conductor to load side grounded conductor will trip it, but a wiggy connection between the same two points won't. Using a Fluke 189 I have 0.004 volts between the line and load grounded conductors...not sure where that voltage comes from.
Don
 
Status
Not open for further replies.
Top