Arcs and GFI tripping

[FionaZuppa]

a friend of mine does all sorts of stuff for halloween and had some bulb flashers made from FS2 starters (a common way to make a bulb flicker). this one was a small 15w incandescent bulb. oddly it would trip gfi recept every so often. so i went over to take a look and i could not find anything wrong in terns of ground faults anywhere, and this flasher was the only added item to the ckt via extension cord. the flasher itself is housed in a 1/2" pull box and is two wire only.

so why would a gfi recept trip on a hot-to-hot arc (thats what the FS2 is doing).

 

[GoldDigger]

The high harmonic content at switching can lead to higher than normal capacitive coupling current between circuit conductors and EGC within the extension cord.
That high current, even though only a fraction of a cycle, could trip the GF circuit.

Sent from my XT1585 using Tapatalk

 

[FionaZuppa]

you mean the capacitive coupling current?
so with a mA meter inline with egc i should see ~5+ mA ??

 

[GoldDigger]

you mean the capacitive coupling current?
so with a mA meter inline with egc i should see ~5+ mA ??

Not really. The meter will be reading a time average, which will correspond mainly to the constant 60Hz fundamental current.
The sub-cycle long higher frequency bursts will average out close to zero over a full period but may still trip the electronics of the GFCI.
To actually measure that current's peak would require a scope.

Sent from my XT1585 using Tapatalk

 

[FionaZuppa]

i have a scope, but will do some other testing 1st, like truing to kill the harmonics with a ferrite choke.

 

[FionaZuppa]

i grabbed his flicker box and ran some tests.
on my garage gfi it would trip every 30-60sec.

here's a crude test i did.

i took hot leg and made coil wrap onto 1/2" rebar using #14 thhn, about 20 turns & 2" long (10 on top of 10) then put some tape over it. this did not stop gfi from tripping. so i then did the same with the neutral on same rebar about 20 turns & 2" long (10 on top of 10) then put some tape over it. so basically two iron core coils on the same iron.

hmmm, the gfi has now stopped tripping.

however, i suspect not capacitive coupling to egc in the cord, but rather something wonky going on in the gfi unit.

 

[ptonsparky]

i grabbed his flicker box and ran some tests.
on my garage gfi it would trip every 30-60sec.

here's a crude test i did.

i took hot leg and made coil wrap onto 1/2" rebar using #14 thhn, about 20 turns & 2" long (10 on top of 10) then put some tape over it. this did not stop gfi from tripping. so i then did the same with the neutral on same rebar about 20 turns & 2" long (10 on top of 10) then put some tape over it. so basically two iron core coils on the same iron.

hmmm, the gfi has now stopped tripping.

however, i suspect not capacitive coupling to egc in the cord, but rather something wonky going on in the gfi unit.

Were they wound the same direction?

Wonky...I like the professional term. Is that close to ‘weird s...?

 

[FionaZuppa]

yes, physically wound the same direction, but because amps flow in one way and back out the other at any given time, magnetically opposite.

wonky, as in the internals are doing something oddly that they probably should not be doing, perhaps clipping in the op-amp and it causes a trip, or a resonance is there causing something to saturate and gfi to trip, wonky like that.