Lights of America does not play well with GFCI

[dacflyer]

i wired up a friends green house this weekend and he is using shop lights ( Lights of America )
since its a green house i wired it up with all weather resistant hardware. including GFCI outlet.
all the lights are the plug in cheap shop lights. so the green house has all outlets.
all lights are brand new - all hardware is brand new.
why do the lights trip the GFCI ? and how to fix it ?

[gerty]

Are the ballasts magnetic or electronic? I ran into a problem with using too many magnetic ballasts on an inverter. I ended up having to break up the circuit so that only half of them came on with each switch. Even though the total load was well within the specs of the inverter, I couldn't turn them all on at once. You can try unplugging all but one and keep plugging in and turning on untill you find the threshold. GFCI's act very similliar with a inductive load.

[rshayes]

If the ballasts were recently made, they are probably electronic ballasts. There seems to be a push to eliminate the magnetic ballasts for energy conservation reasons.

The electronic ballasts are basically switching regulators arranged to give a constant current output. The switching is done at a high frequency to reduce the size and cost of the magnetic components. The sharp edges on the switching waveforms have harmonics that can interfere with radio reception. Most of these devices require radio frequency interference (RFI) filters to meet FCC rules. These filters typically have a line to neutral capacitor and smaller capacitors from line and neutral to ground. These capacitors pass small currents (in the milliamp range) with 110 VAC across them. The line to ground capacitor will have 110 VAC across it while the neutral to ground capacitor will (or should have) only a low voltage across it, since the neutral and ground are usually connected together at the service entrance to the building (by current electrical codes). The result is that a small current flows from the line to ground.

The ground fault interrupter (GFI) operates by comparing the currents in the line and neutral. If the currents are not equal, the GFI disconnects the load.

The current from line to ground through the RFI filters does not pass back down the neutral conductor. This causes the line and neutral currents to be unequal from the point of view of the GFI. Each RFI filter added in parallel increases the inequality of the currents and eventually the unbalanced current will be sufficient to trip the GFI. I haven't dealt directly with the florescent ballasts but I have had this problem when using several switching supplies on a circuit with a GFI.

I don't know of a good solution to this problem. If you remove the capacitors, the filters will not work properly. Alternately, if the ground current is also subtracted from the line current, the ability of the GFI to detect certain ground faults (line to metal case, for example) is lost.

[dacflyer]

yes, they are electronic ballast bout the size of a pack of cigarettes and they operate 2 4' lamps T12 or T8
my only solution is to either use ground isolator blocks
( them orange or grey plug adapters to convert 2 prong to 3 prong ) or just not use a GFCI outlet at all.
but its in a small green house and water does get on the floor, all outlets are 3' off the ground and 1 on the ceiling.
the outlet all have the weather guard covers over them. (bubble covers) i really do not see a problem, but i wanted to stay to code..
i will try to call L.O.A. again today..

[haklesup]

were the power cords pre wired on those cheap shop lights or did they leave that step to the installer?

Reversal of the hot and neutral conductors may cause this. Reverals of ground and neutral would also do thas but its unlikely you would make that kind of mistake. reversal of ground and hot would probably trip the breaker for overcurrent and not GFCI action, the difference is indistinguishable.

Check the shop light plug with a DMM and visually inspect for proper wiring. ALso double check the proper wiring of hot, gnd and neutral leading up to the greenhouse. Maybe somebody played games and connected neutral to ground at a sub panel instead of the main panel. You can also consider trading the GFCI outlet for a GFCI breaker in the panel instead.

How does that GFCI plug work with a receptacle tester (three prong thing with lights and a trip button). You should use one of those to check the proper wiring.

Another gotcha with GFCI is sometimes the downstream sockets are wired in parallel to the GFCI input when they should be wired to the GFCI output terminals, reversal of hot and neitral is possible here as well as at red wire hots for 3 way switches.

My gut geeling is you have a miswire somewhere

[dacflyer]

i did all the wiring myself, all is correct, the outlet tester shows all to be correct.

the light fixtures all came with prewired cord attached.
i can take the light and try it on any other gfci outlet in the house and it trips also..
i opened a fixture, and they are also wired correctly.
L.O.A. has still not responded. Grrrrr..

[Bob Scott]

i did all the wiring myself, all is correct, the outlet tester shows all to be correct.

the light fixtures all came with prewired cord attached.
i can take the light and try it on any other gfci outlet in the house and it trips also..
i opened a fixture, and they are also wired correctly.
L.O.A. has still not responded. Grrrrr..

Things to check:
Check for a neutral to ground short somewhere in your wiring - a wiring fault after the GFCI and before the lights. An outlet tester cannot detect this type of wiring fault. But if it trips your other GFCI's, it's not this fault.

The lights themselves may be diverting some current to ground. If there is a capacitor to ground in the light ballast, it might trip the GFCI. It would have to be bigger than about 0.6uF in order to divert 30mA according to my quick calculation. (Xc=V/I, Xc=120/.030 = 4000)(C=1/[2*pi*60*4000])

I vote that your new lights are at fault.
Check to see what current level your GFCI's trip at. It might differ from 30mA.

[dacflyer]

oh i know its the lights as the problem, others on the net have same problems with these cheap shop lights...
i am gonna try to call L.O.A. in a few min...



ok, update, i just got in touch with Yoshi Okamoto ( bad english chineese guy )
he says they test the lights extensively in lab and have no such problems.
(figures ) i told him there is a lot of post on the net about such problems.. but they deny it.. so i guess i will not call the MFG of the GFCI outlet and see what they deny also..

to be continued......

[CeaSaR]

Have you thought of putting the lights on 1 circuit with 1 of those PIR auto switches
(high enough out of the way so that no one can touch it), and then putting a switch
on the circuit with the GFCI that runs an optoisolated flipflop to control the manual
on/off pushbutton on the PIR auto switch? Keeps the normal everyday hand operation
safe while letting the cheap-a** lights function "perfectly" . Yeah, you'd have to
hit the breaker to change a light bulb, but at least they'd work.

Just a thought...

CeaSaR

[gerty]

Have you tried just one fixture ? How many total do you have?

[dacflyer]

PIR will not work due to the lights will be on all the time for the plants..

and yes i have tried each light separately. just to eliminate any possibility of a fixture being the culprit.
all six of them do it..
as an experiment i am going to try 2 other name brands of GFCI's and see what results i get.
stay tuned

[haklesup]

It trips when you plug it in, can you then reset it while its still plugged in or does it stay tripped? If you can reset it, it may be inrush current from somewhere.

To get to the bottom of it, you could rig up an extension cord with ammeters on the hot, neutral and ground seperately and measure the current on each leg directly to see if that leads to an explanation. However the event may be too fast to see with a DMM and may need to use a scope

[dacflyer]

nope, cannot reset if its still plugged in...already tried that..
i am sure its gonna be too fast to measure any current in rush. and the outlet is rated for 15 amps.
i highly doubt in rush is more than 15 amps, fixture has built in 2 amp fuse.

[reloadron]

Hi Ya Dac

I may be way off base with this but my understanding is that some fluorescent lighting using the newer ballast designs that incorporate capacitors tend to trip GFIC outlets and breakers. The logic seems to be that at turn on instant the current in is not equal to the current out and that short duration, while the caps charge, causes tripping. That or somewhere in the scheme there is leakage to ground.

Ron

[dacflyer]

i do not know , but i am trying (testing ) out 2 other GFCI's (different brands) tomorrow and see what results i get..