GFCI Nuisance Trip Physics

[wwhitney]

Say I have a 2 wire 120V circuit going to a piece of utilization equipment, and say there's no connection between the equipment and earth or any bonded metal (in particular no EGC).

1) Can a non-faulty GFCI nuisance trip in this situation, and what is the physics by which this happens?

2) Kirchhoff's current law is instantaneously obeyed at all points in time, yes? There aren't circuit elements that can say accumulate electrons during one part of the cycle and release them during another to actually introduce a difference in the currents in the two wires, right?

3) So then I infer that certain aspects of the current waveform through the CT in the GFCI are able to induce a signal in the control circuitry even though the difference in the currents is always zero? E.g. high frequency harmonics?

Thanks,
Wayne

 

[David Castor]

The toroidal current sensor inside the GFCI senses net flux between the neutral and line. Since the net current is normally zero, the net flux is zero. But this sensing system isn't perfect. There is always small amounts of stray flux that isn't perfectly coupled into the sensor. For loads with high transient current, there can be enough flux summation error to generate a trip signal. On industrial systems, we often see a similar issue with the flux summation CTs used for ground fault detection. If the relay is set too low, it will sometimes trip during transformer inrush or motor starting.

Capacitive coupling to ground on long circuits can also result in tripping, although that is a different issue. This is more commonly seen where the GFCI function is in the circuit breaker at the panelboard.

I suppose EMI, either conducted or radiated, could also cause misoperation by overwhelming the electronic circuitry inside the GFCI. I don't think I've have seen this in real life, but seems possible.

I recently had to replace a kitchen GFCI that wouldn't reset. The local store had an off-brand unit made in China for $15. They also had an Eaton for $30. I bought the Eaton.

 

[junkhound]

Common mode EMI from power supply converters can trip GFCI even when 60Hz currents are perfectly balanced.

common mode current occur when the switching frequency couples thru the FET or IGBT drain or collector capacitance to the heat sink.

I had to add a 10uH inductor in the green wire to chassis on a 1 kW, 120kHz 3kV supply I built to prevent GFCI trips.

If your utililization equipment or leads downstream of GFCI have any leakage to earth, easy trip
e.g plug and socket of extension cord in the rain laying on the ground, water pump with leakage path, etc..

 

[junkhound]

Common mode EMI from power supply converters can trip GFCI even when 60Hz currents are perfectly balanced.

common mode current occur when the switching frequency couples thru the FET or IGBT drain or collector capacitance to the heat sink.

I had to add a 10uH inductor in the green wire to chassis on a 1 kW, 120kHz 3kV supply I built to prevent GFCI trips.

If your utililization equipment or leads downstream of GFCI have any leakage to earth, easy trip
e.g plug and socket of extension cord in the rain laying on the ground, water pump with leakage path, etc..

 

[junkhound]

another example would be an old 'all american five'* radio chassis sitting on a wet concrete counter or similar - those old radios often had up to a 0.2 uF cap to chassis from the line, > 5 mA, plugs then seldom were polarized, so ground current possible depending on which way plug inserted. .

My very first electric shock experience (just a tingle) was sitting on the furnace duct while operating phonograph in the early 1950s.

* if you dont remember DC distribution in NYC, ask your grandpa