Hello everyone, been lurking for a while, now have a question. Has anyone had any success running HID lights, specificaly HPS light fixtures on a GFCI breaker? Put power to a boat dock, installed 2 HPS fixtures, and cannot get the breaker to reset. These are new fixtures, and will function normally when plugged in on a non GFCI circuit. I can un-hook the neutral wire going to these fixtures and the breaker will reset, allowing the rest of the things on the circuit to function normally. Only thing I can figure is a neutral to ground bond in the ballast causing the breaker to trip. Does anyone know of a manufacturer that has lights that will work on these circuits? The fixtures in question are made by Lithonia. Any insight would be appreciated.
Simply operating an HID light on a circuit protected by a GFCI should not be a problem. However ...
Ballasts (and HID fixtures have ballasts) often fault to ground. That will trip the GFCI. The solution is to replace the ballast.
As for breakers marked "HID," here is what UL has to say: "A circuit breaker rated 50 A maximum, 480 V or less, and intended to switch high
intensity discharge (HID) lighting loads on a regular basis is marked “HID.” "
Reno,
The fixtures are new, I took them out of the box, hard for me to believe both ballasts are bad. Could it be a manufacturing problem. Both will work on a non-gfci circuit.
Jimmy:
1: Are the ballasts multi-tap (120-240V)??
2: Can you make the circuit 240 volt??
3: Are the fixtures and wiring 'out of reach'??
John,
They are not multi tap, but they are out of reach. Thought about putting them on a non Gfci circuit, but worried about some one tapping it for other uses.
Jimmy;
Yes, I know it's tough to idiot proof your work once you are gone.
I thought that most of the new fixtures came with multi-volt ballasts, but you just killed that.
Just a thought, did you try the circuit with the ballast ground conductor disconnected? If it still trips, and the ballast is leaking to ground, then you could try putting fiber/plastic washers to isolate the ballast case from the fixture body. (Internal leakage to ground). Keep the fixture body bonded at all times. You could then contact Lithonia to determine if they will replace the ballasts.
Just a thought...
I will check that, thanks for the insight John.
update: I went with Johns advice, and used some rubber to isolate the ballast from the fixture housing, and bingo, the breaker will reset and hold. Took the rubber out just to verify and the breaker will not set. Waiting on an answer from Lithonia to see if they are going to replace them or just ignore the problem. I could not see a bond from neutral to ground, nor could I read any leakage with my Fluke meter, but the proof is in the pudding. I will bet Lithonia knows the problem exists just havent had enough complaints to do anything about it yet. HID lighting on docks for fish attractant is just starting to catch on in my area, bet there will be more trouble to come.
Jimmy:
If you don't get anywhere with the mfg., you did find a "fix". IF you have a ballast 'in stock' you may want to swap it out to see IF that is the 100% cure.
Perhaps Greg has some input on ballast leakage to ground?
I can say I never came upon this issue in my 'tool days', and the majority of the HID lighting I worked on was 208 to 480 volt.
I suppose the first thing would be to see if it is the hot or neutral leaking (what voltage do you measure from the floating ballast to ground?) but it really doesn't matter. Both represent a potential hazard.
I am curious what the mfg says.
It is possible that this is just the RF filter providing a fault path. That used to be a problem on some old computer power supplies until U/L tightened up the listing standard.
Greg, off the cuff I am going to say neutral leakage, because the breaker would not reset with the light hooked up and the switch turned off. I did not measure voltage from ballast to case after isolating them.
John, I did try with the ground unhooked but figured it would not help as the dock is all aluminum and of course bonded at every possible location, and the lights are mounted on aluminum boxes. I will update when/if Lithonia replies and what the solution is.
I'll bet that Greg is right and that they got sloppy on their RF bypassing. If you hook it up non-gfci, and measure the ground current with your Fluke on AC mA, you will probably observe enough current for a legit gfci trip.
Joe