Questions with regard to GFCI's and AFCI's and how they work/test.

GFCI’s operate by measuring and monitoring the current imbalance between the grounded and ungrounded conductors. This is done by a current transformer built into the device, either a receptacle or circuit breaker. At .5 mA the circuit protection device opens. However, the equipment ground conductor serves no functional purpose with regard to the ground fault protection circuit. Why then will my GFCI tester not trip a GFCI receptacle device if the receptacle is ungrounded? I have experienced this during inspections. I have also bench tested this to prove the theory. Additionally, the NEC permits replacement of non grounded two pole receptacles with GFCI receptacles provided the GFCI receptacle is marked with “No Equipment Ground”. All downstream receptacles protected by this device (4 max) must be marked “Protected by GFCI” & “No equipment Ground”. If a non grounded GFCI receptacle with not trip using a GFCI tester how is it protecting ungrounded downstream devices?

I’m less familiar with AFCI’s. Do they operate on the same principle as the GFCI. That is a current transformer measuring and monitoring the current imbalance between the grounded and ungrounded conductors. I would assume this device is looking for current spikes. Anyone understand?

When you push the button on your GFCI tester it produces a simulated fault current from the “hot” lead to to the “ground pin” of the tester. Because you do not hav a “ground” available no current flows to this path and the GFCI does not detect any fault.

use the SEARCH feature and put in the terms " GFCI " or " AFCI " and you will see we have covered this many times and many educational information has been posted many times regarding these devices.

Here is a discription of how AFCIs work.

Okay, I did the search and there are many, many post. Non of which seem to answer the qustion.

The GFCI device will open when there is a difference between the ungrounded (hot) and the grounded (neutral) of 5 milliamps.
The tester that you are using to achieve the results connects the ungrounded (hot) conductor with the equipment grounding conductor and if there is none available then it will not work.
According to UL these testers are not to be used when testing GFCI circuits. The test button on the device itself is the only true test.
When finding if a receptacle is on a GFCI circuit one of the testers can come handy if there is an equipment grounding conductor present. If there is no equipment grounding conductor present then the tester is useless.
If checking to see if the GFCI will trip at the recommended current the little tester is useless.
To make a report and state that the GFCI circuit is working correctly after testing with the little tester would be a false statement as there would be nothing to show at what level of current the device opened. The only comment that could be truly made is that the receptacle is on the same circuit as the GFCI device.
Arc Fault works in the same way as GFCI but at a higher level of current. The tester will not impose an arc fault in the circuit so again all that could be stated is that the receptacle is on the same circuit as the arc fault not that the arc fault is working properly. These testers are only good for identifying the circuit that a device is on but are worthless for anything else although a lot of misinformed people will swear by them.

The answer to your question is in post #2.

The GFCI tester’s test button, when depressed introduces a high resistance ground across the ungrounded conductor and earth ground. This high resistance ground creates a current imbalance measured at 5 mA. Otherwise a bolted dead short current flowing to earth ground would trip the circuit breaker protecting the circuit from overcurrents and short circuits, Regardless where in the circuit in question the GFCI tester is placed, this 5 mA current to ground by way of the earth ground conductor creates the imbalance between the ungrounded and grounded conductors throughout the circuit. The continuously monitoring current transformer within the device recognizes this current imbalance and opens the circuit.

The tester’s test button can not be paralleled across the ungrounded and grounded conductors or the device would see it as a load and the current would remain balanced. Therefore no fault would be detected.

With the absence of an earth ground the tester’s test button has no path to ground when depressed to introduce this fault current. Therefore no method of providing a true and accurate test of the GFCI device and/or circuit. However, with the proper wiring in place including the proper earth grounding conductor, based on the controls stated above I would continue to suggest a positive test would indicate the GFCI device is functioning properly. With indication from the tester of an open/missing ground conductor any conclusions with regard to the GFCI device would be in error. This lack of earth ground should be indicated on the report however with a recommendation to correct.

The earth ground has no connection or impact on the functionally of the GFCI device only on the ability to test.

AFCI’s continuously monitor currents and interruptions/fluctuations to currents. The device distinguishes the deference between arcing faults and natural or normal current variations. They monitor intensity, duration and frequency of the arcing fault,

Although what you have quoted is in essence what happens, it is not completely accurate.

Through the bonding of the equipment grounding conductor and the grounded neutral in the service equipment the conductor that is referred to an “earth ground” is carrying the current back to the source from which it came. With the plug in tester the hot is shorted to the equipment grounding conductor which is bonded to the grounded (neutral) at the service thus grounding it back to the source.

To think of the current being carried to earth is the base for the confusion about the purpose of the equipment grounding conductor.
The earth itself plays no part in the carrying of current and no function in the operation of the GFCI device.

Yeppers…over time more and more people are learning that the EARTH at the dwelling is not serving the functions many think it is…part of what i do is look for misleading statements in publications that list the myths that electricity is trying to get to earth in a fault…oh how WRONG they are…:slight_smile:

GFCI’s work of differences in the balance between the Grounded and Ungrounded conductors…and as Mike states so well…has no effect really on a GOOD return path to the source…now in regards to the function of the OCPD…a different story…:slight_smile:

Alas…I wish they would consider changing the name of the EGC to the EBC…but alas…futile fights are never won by the weak…:slight_smile:

Yeah! What Paul Said. :wink:

Yeah…I agree…What William Said…:wink:

Thanks for the dialog. I love talking about this stuff. Earth ground, equipment ground. Same basic thing. Along with bonding they create the short circuit protection for any electrical system. While as I stated the grounded conductor or neutral provides a return path for current. So there’s really no confusion with regard to grounding. How the topic of grounding entered this conversation was how does the equipment ground effect the testing, and testing only of a GFCI device. It’s now understood that the test button creates a connection between equipment ground and the ungrounded conductor (hot) causing an imbalance between the grounded and ungrounded conductors greater then 5mA. Which trips the GFCI device.
I still am at a lost for how to test the GFCI device accurately and its ability to open the circuit at 5mA. If the little tester will not, does a Suretest Circuit Analyzer? Which cost $350 to $400. I find it hard to believe the only acceptable method to test is by depressing the test button on the GFCI device itself. How would one know if it works and works at 5mA on a fifteen year old device on an outside rear porch in extreme climates?

Is there any way to truly test???

Thanks again for you valuable input.

The test button IS a real test, the only one supported by the manufacturer. It does not depend on the grounding conductor at all. They just create an imbalance across the current transformer in the device.