DLB wrote: ↑Fri Jan 28, 2022 9:02 am
One thing I learned from this video is that the GFCI circuitry is monitoring for downstream "Grounded Neutral" (GN) in addition to Ground Fault (GF). One thing I did not learn, however, is exactly how that GN circuitry works. I have one GFCI circuit in my house that suffers from nuisance trips, aka ghost trips, maybe a rate of 10X per year. I see no straightforward way to distinguish whether the cause is relating to GF or GN, or what sort of transient would cause one or the other on such as infrequent basis.
- David
I am probably going to regret coming out of retirement to post this, but here goes. JPG had some good insights over on this thread (
viewtopic.php?p=286821#p286821). As JPG pointed out he found a high frequency sensitivity of the comparator circuit used in GFCI's used where he worked. In watching the YouTube video posted on this thread about the circuit used in GFCI's, I can now better understand why this nuisance tripping "might" occur. From this video, they state "the fault current through the ground connection creates positive feedback that causes the sense amplifier to oscillate." When the oscillations last longer than the time delay circuit then the SCR fires to disconnect the load from the line. IF there are sufficient harmonics or noise from line connected equipment then it could easily trigger an oscillation in the sense amplifier that gets mis-interpreted as a ground fault (positive feedback triggers it). Also, if you listen to the YouTube video, he points out that the UL test circuit and Fairchild's application note has quite low resistance on the line conductors, so I wonder if less than ideal ground/neutral connections could exacerbate the nuisance trips? Looking into this and/or pursuing line filtering for a GFCI that nuisance trips is where I would start looking if I had this problem. The video also talks about some GFCI's having a 2nd toroid (sense inductor) that senses the current between ground and neutral so it might be nuisance trips are through this path so another reason to take a hard look at neutral and ground wiring and bonding if nuisance trips are common IMHO. Just my 1 cents (further reduced due to inflation at the moment
).
It seems like there is a myth on this thread that GFCI's are all protecting which is being busted. I agree that GFCI's don't protect against everything and that was never their intent. Even with a properly installed and functioning GFCI it is still possible to be shocked. There is a phenomenon when someone gets shocked by AC electricity where the muscles contract and can't respond fast enough between AC cycles to release and separate from a live powerline. GFCI's were only intended to break that connection should a ground fault happen, thus giving the person a chance to remove their hands from the live wire and prevent being continually shocked. The specs for GFCI's are designed around this intent based on preventing muscles from seizing during a shocking event. I don't know how they did these tests, but there were human tests performed that showed muscles tend to seize at around 10mA of fault current so it is believed that at 10mA and above that a person can't let go of a live AC wire. This is why the GFCI limit is typically set at 6mA (well below the believed point of no return). So, in a normal application with a properly functioning GFCI with a plugged in device that faults, can a person still be shocked? I would say it is possible, but not likely. Even if they are shocked, the GFCI should respond before the fault current gets too large that they can't remove their hand. Can they be electrocuted? I highly doubt it, but then again freak accidents are exactly that...freak accidents. Bottomline is GFCI's were NOT intended to prevent any and all shocks, but rather were designed to prevent muscle latching whereby a shock leads to electrocution because you can't let go of the live wire.
What I don't understand are the schematic drawings on this thread. As has already been pointed out, there are direct connections between line and load in the drawing and a GFCI would NEVER be connected this way so why are there direct connections there? This removes the GFCI from the circuit because it bypasses it, so I don't understand the intention behind this schematic drawing. Also confusing for me is why the human was drawn in a series connection in the other schematic. Clearly this wouldn't be the intended user (human) of the GFCI, so it would have to be an electrician working on the circuit in order to put themselves into the circuit path, but then why are they working on the circuit live to begin with?