0

I have a GFCI receptacle that takes 2-5 seconds to trip when testing with a standard 3 prong receptacle tester. The test button on the receptacle trips right away but two of my trustworthy handheld testers (different brands) take much longer. After changing out the GFCI, the circuit can be tripped immediately with the testers.

Is this a common failure mode? I've replaced many that trip unnecessarily or won't reset at all, but this slow-trip is a first for me. (and has now been replaced)

This also leads me to wonder if there is a tester that measures the time to trip. I'm unlikely to notice the difference between 30ms and 300ms when punching the test button, yet this could be a serious hazard with the equipment, right?

5
  • "test button on the receptacle trips right away" - then it's fine. Your "trustworthy handheld testers" are for finding open neutrals, missing grounds, and reversed polarity.
    – Mazura
    Oct 14, 2019 at 21:15
  • Are you saying that GFCI testers can't test GFCIs? Oct 15, 2019 at 1:22
  • I'm saying I don't use them to test the functionality of a GFCI receptacle. That's what the integral 'test' button is for.
    – Mazura
    Oct 15, 2019 at 1:27
  • I've had to replace so many bad GFCIs that I just don't have a lot of trust for them. This one is new behavior in my experience, and when I effectively connect hot to ground and the GFCI doesn't trip, something seems wrong. Oct 15, 2019 at 2:20
  • Agreed. That's why after "the problem went away with a new GFCI" I stop asking questions. Do note though, a GFCI is looking for an imbalance on the neutral, while your tester likely shunts to ground. I suspect you're missing the No EGC sticker and there isn't one (a grounding conductor).
    – Mazura
    Oct 15, 2019 at 2:38

3 Answers 3

5

While this is an older question, it is a top Google result when I was researching the same issue and my findings are different/conflict with the previous answers provided above so thought I would add if others come across this same issue.

UL943 is the spec for GFCI's. It does specify maximum time for a GFCI to open as

T = (20/I)^1.43

where I is in mA. From that equation, 20mA can take 1 second, 15mA can take 1.5 seconds, and 6mA can take 5.6 seconds.

I checked my tester and it has a 18 kOhm resister for 6.7mA so it can take almost 5 seconds to trip per code.

2
  • I am shocked that the time permitted is so long. But I managed to find a copy of 943 and confirm it. Sep 6 at 15:30
  • Thanks for the details. I do recall reading something like that somewhere later in my research but that's an excellent and concise description. I regret that I trashed the old receptacle and never had the chance to definitively resolve that mystery. Sep 6 at 22:38
2

It may be a faulty tester, or a faulty ground. The tester relies on a wired ground to simulate the ground fault. In fact they won't work on ungrounded connections, and that's not a defect.

4
  • So, if my two testers successfully trip every other GFCI, then I should suspect a faulty ground. Since the problem went away with a new GFCI, the faulty ground was within the old GFCI receptacle? What kind of faulty ground would cause this? Not an open ground, right? Oct 15, 2019 at 1:22
  • @user7264855 I did not have the information that you had replaced the receptacle and that solved the problem. My answer reflects that knowledge gap. Oct 15, 2019 at 2:59
  • the ground issue could have been between the ground wire from the junction box and the ground screw on the outlet. Id' be curious if a re-wired original outlet would still exhibit the issue. Oct 18, 2019 at 14:58
  • @PaulBelanger for that matter the loose ground could be anywhere between the outlet and the neutral-ground bond in the panel (or pole if the panel one is out). GFCI devices don't communicate with ground in any way. Indeed I would suspect the N-G bond. Oct 18, 2019 at 16:05
-1

Trip time is not part of the specs for GFCIs or the test circuits, only the level, which is between 4 and 6mA of current that is NOT returning on the neutral. It could be that your handy tester is using a current limiting resistor inside to test the circuit that takes 3-5 seconds for the trip threshold to exceed that value. The test button on the GFCI basically works the same way (a current limiting resistor) but probably uses a lower ohm value so that it reaches the threshold faster. Not something to worry about really.

5
  • Well in my opinion it is because you are slow. I have tested thousands of breakers and , outlets ! There is a specific fault level IF those specific values are met , the protection opens the circuit and you are safe.
    – Ed Beal
    Oct 15, 2019 at 0:33
  • @JRaef That's good info. I measure the resistance in one of my testers at 15.6 k Ohms and the other at 1.6 M Ohms (hot to ground when button pressed), which would put the first at 8 mA and the second at about 80 uA. (if I haven't screwed up my math). Now I don't understand why the 1.6 M Ohm version (a Klein) trips at all; but it does with every other GFCI. I don't understand what you mean by taking time to trip the threshold value of 4-6mA. Current (Amps) is a rate, not an amount of flow (which is Coulombs) right? Does the resistance in a GFCI vary over time or with voltage? Oct 15, 2019 at 1:54
  • 3
    Actually trip time is part of the spec for GFCI, The trip time should be within a few AC cycles, much less than a second. Oct 15, 2019 at 11:55
  • Resistance in the tester may be from an NTC resistor, meaning the resistance drops as it heats up. Those are common to use as inrush current limiting devices.
    – JRaef
    Oct 15, 2019 at 17:28
  • "Trip time is not part of the specs for GFCIs" - it absolutely is. UL943. Obviously it must be - if it weren't, a GFCI could take, say, an hour to trip, which wouldn't be very useful. Even if the spec were for a something that translates to "almost instantly" (like "within one AC cycle") that would translate directly to a specific time duration.
    – nobody
    Sep 5 at 15:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.