Testing a GFCI receptacle

Status
Not open for further replies.
Little Bill

Little Bill

Moderator
Staff member
Location
Tennessee NEC:2017
Occupation
Semi-Retired Electrician
I know the test button on the receptacle is the correct way to test one. I also know that if using a plug-in tester it will trip/work if there is an EGC connected, but won't if no EGC. However, I have found this twice in the last few months where a receptacle had a properly connected EGC but a plug-in tester would not trip the GFCI but the test button would. The customers ask me if the GFCI needed replacing. I tell them the test button should be the defining answer. But I'm wondering why a plug-in tester will not trip the GFCI when it has an EGC that is properly connected. These have both been in houses less than 20 yrs old.
 
The obvious answers are either that the receptacle's EGC pathway is not intact, or tester is defective.

I use my solenoid tester as a GFCI tester. I can test to nearby grounds if the receptacle EGC doesn't trip it.

Did you try the tester in the GFCI itself? Do you know for sure the EGC pathway is intact?
 
There is also a chance that the current drawn by that particular tester is low and the GFCI trip point is at the high end of the range.
I assume you were using the same tester both times?
You can measure the resistance from hot to EGC on the tester with the button pushed. (Not plugged in, of course.)
 
I did check for bootlegged grounds and found none. I didn't check all the way to the panel but the plug-in tester showed proper ground. Also, I used two different plug-in testers so I'm pretty sure it wasn't my testers.
 
210619-1503 EDT

I believe the test resistor in a GFCI is about 15,000 ohms. At 120 V this will prove a load current of 0.008 A (8 mA) from hot to neutral or EGC.

So you could try an external 15,000 ohm resistor and see if it will trip the GFCI. Power dissipation is almost 1 W so I would use a 2 W wire wound resistor. You would connect from the receptacle hot output to EGC. You can not use the receptacle neutral unless you can get to the neutral coming into the receptacle.

.
 
Little Bill said:
I did check for bootlegged grounds and found none. I didn't check all the way to the panel but the plug-in tester showed proper ground. Also, I used two different plug-in testers so I'm pretty sure it wasn't my testers.
Click to expand...
I have had this happen on a old rag wire circuit where grounding was done via the undersized 16awg ground wire that was simply folded back on the cable and box cable clamp making the ground. The simple plug in tester would show a ground but the meter showed a fairly high resistance and the GFCI tester wouldn't trip the GFCI. How much of a safety is this was I'm not sure, as it is claimed that these testers are not an accurate test means.
 
Fred B said:
How much of a safety is this was I'm not sure, as it is claimed that these testers are not an accurate test means.
Click to expand...
None, really, as a GFCI depend on current to earth (because we use grounded supplies), not an EGC.
 
My friend had the same issue and he found out that the newer devices don't work with plug in testers. It was either gfci receptacles or CH gfci breaker I am not sure
 
If the plug in tester does not work and the device push button does, then the plug in tester is not flowing enough current for the GFCI to trip. Maybe the resistor in the plug in tester is not sized the same as the one in the device, or as others have suggested there is not a good EGC path at the device.
The built-in resistor is required to be sized like this.
5.15.6 The current imbalance employed by the supervisory circuit shall be sufficient to cause tripping at
85 percent of rated voltage, provided that at rated voltage the current imbalance shall not exceed 9 mA.
Click to expand...
 
The scenarios mentioned above are the most likely causes for the failure of plug-in tester to work. Below is a less likely but still possible explanation:

Are there one or more downstream receptacles that are fed by the load side of the GFCI receptacle? If so, then there might be a bootleg ground connecting N and EGC downstream of the GFCI, and it was put there because of a break in the EGC path that's somewhere upstream of the GFCI receptacle. In this case the current from the plug-in tester when the button is pushed would return on the load neutral of the GFCI instead of through the EGC back to the panel, and so it would not trip. But the plug-in tester would show that the wiring is correct because of the bootleg ground that connects the downstream N and EGC.

To check if this is could be happening, disconnect the load side neutral wire on the GFCI receptacle (assuming there is one). If the plug-in tester now indicates a bad ground then what I described above is probably what you have.
 
Last edited:
Fred B said:
I have had this happen on a old rag wire circuit where grounding was done via the undersized 16awg ground wire that was simply folded back on the cable and box cable clamp making the ground. The simple plug in tester would show a ground but the meter showed a fairly high resistance and the GFCI tester wouldn't trip the GFCI. How much of a safety is this was I'm not sure, as it is claimed that these testers are not an accurate test means.
Click to expand...
so maybe this fits here>
www.sparkyfacts.co.uk

Earth Loop Impedance Test - The Illustrated Guide | SparkyFacts.co.uk

Insulation Resistance Test explained and illustrated with pictures. Guide for the inspection and testing earth fault loop test exam.
www.sparkyfacts.co.uk www.sparkyfacts.co.uk
~RJ~
 
Another possibility is that at low ground fault currents the time to trip can be several seconds.

A class A GFCI with a trip current threshold of 6mA can take almost 6 seconds to trip and still be compliant with UL standards.

Jon
 
I should have said in my OP that in the 1st case I replaced the GFCIs as the customer requested it as the house was being sold. The replacement GFCI did trip with my plug-in tester. The 2nd case, I didn't replace it.
 
Status
Not open for further replies.
Top