Today at work some hid lights needed to be disconnected from one generator and reconnected to another. The lights were unplugged from a standard 15A receptacle and plugged into a 15A GFCI receptacle. When turning on the lights they tripped the GFCI plug at the new generator.
Nothing had changed from the time of disconnect and reconnect. Out of curiosity I took a reading across the hot and ground prongs on the cord end. I got an initial reading of about 700,000 ohms, but that number immediately began to slowly fall. I watched for about a minute as the resistance fell to about 500,000 ohms.
It would have been nice to leave the meter on for a while to see where the number ended up, but some people wanted light so they could get back to work. Seeing that the ground leakage appeared to be minimal, I figured that this must be an inherent property of hid lighting. I removed one GFCI plug and replaced it with a standard receptacle, plugged the lights back in, and they work fine now.
Why the ground fault? Or is it even a fault? Something to do with the capacitors? The coils?
It seems to me that there should be zero continuity to ground.
Hopefully somebody can put my curiosity to rest with a good explanation because it isn't making sense to me.
If there nothing wrong with the circuit feeding the GFCI, 99.9% of the time the GFCI dedected leakage. The light should have not tripped the GFCI. Exactly where is anyone's guess. Possibly in the transformer would be my first guess. The only way to know is to disassemble it and meg the windings .
I consider 500,000 ohms as unaceptable. Get a megger on it and find out what you really have. Meters are not real good indicators of insulation breakdown. But do excellent on dead shorts. There is much discussion in our industry concerning how much leakage is to much. The number seems to vary from person to person. For example: for me, motors reading less than 5 meg ohms it gets pulled. Exception would be DC motors where you can really not have a zone of confidence.
Discharge lighting will randomly trip 6ma GFCI devices - be it Fluorescent or HID. Must be the Lamp(s) charge state "falling" back through the Grounded Conductor, which comprises the imbalanced load, and ultimately trips the GFCI (that's the only thing I can see causing the random trip issues).
As to the continuity between Line and Equipment Ground, there should be an extremely high Resistance value - above 10 Meg Ohm. Should find a very low Resistance between L-N, unless the Ballast is a straight Linear Reactor without an Autotransformer section, in which case there will be a very high L-N Resistance; not >10 MOhm, but above 100KOhm.
In your case, maybe a fixture has a pinched lead grounding out. That would definitely trip the GFCI, as some level of current will be flowing on the equipment Grounding Conductor - resulting in an imbalance between L-N at the GFCI device.
The way you describe the circuit charging up (went from 500KOhm to 700KOhm), sounds like a Capacitor is being charged in at least one of the connected fixtures.
If these are HPS (High Pressure Sodium) or PSMH (Pulse Start Metal Halide) type Lamps, then the Ignitor may be what is charging between Line and Eq. Ground.
If you plan to Megger the fixtures, only megger the Ballast's Windings. Disconnect the Capacitor and the Ignitor (if used). Also remove the Lamp from the socket, and disconnect the input branch circuit (L+N) before testing with the megger.
Connect one of the megger's leads to the Ballast Winding's input termination; and the other megger lead to the fixture's housing.
A quick and dirty method to see if your Ohm Meter was seeing a Capacitor is to place the Ohm Meter on a lower setting - something in the 100K range; then take a reading like you did before. 1: First discharge the circuitry by physically shorting the Cord Cap's blades from L-N, L-G and N-G (use a piece of steel stud as a shorting device!!!) 2: Connect the Meter's Leads, and verify the "charging characteristics" explained below: *a: Reading should begin at a low number, like 100 - 500 ohms, *b: Resistance should rapidly increase, *c: Resistnace should taper off at around 100K, or display "Infinity"
If the Meter's readings do the opposite, you are most likely reading an Ignitor.
Scott " 35 " Thompson Just Say NO To Green Eggs And Ham!
Scott, your comment concerns me - and I was not aware of this 'leakage.'
I am concerned, as, in the latest edition of the NEC, essentially every household circuit will have some sort of ground fault protection on it. Combine that with the recent proliferation of 'energy' codes (and their bias in favor of fluorescent lighting), and I see a problem being created.
Thanks for the help Scott. My meter is auto-ranging, but I'll try metering the cord end as you suggested. If this is inconclusive then my chances of figuring out exactly what is happening with the lights are not good. This is my last week working with this company and my next few days will be busy with roughing in the offices and showing the new electrician what has happened so far on this job. I just won't have the time to go get the megger and do a thorough investigation.
As far as the lights, they are 400W M.H., pulse start or not I don't know.
A large part of the problem is probably the cables feeding these lights. They look like they have seen better days. The lights are temporary for construction and have been used many times...and so have the NMD cables powering the lights. Maybe I'll run an extension cord up to the lights and see if the GFCI trips then.
I explained what was happening to the new guy and may just have to leave it at that if I can't make the time to check this out some more.
I am not 100% sure on the whole leakage thing, but it's the only thing that I can think of which would trip a 6MA GFCI.
Had some Fluorescent jobsite temp lights that I made out of 4 Foot 2 Lamp strips, attached to some clean stands (made from 3/4" EMT). Each stand had a GFCI Receptacle, which fed the lights on the Load side.
When plugging them in (at the spider box), random device trips would occur: * Never the same fixture stand. * Never all at once. * Always on initial turn on. * Most of the time there was no trip.
It was very unpredictable, and at times very frustrating (seemed like the ones in the darkest areas would trip the most). No leakage during operation - only leakage at start, which leads me to think the high electrostatic field which surrounds a discharge Lamp at initial start, was causing the devices to trip.
Similar to a Rapid Start Fluorescent Lamp experiencing starting issues, where the fixture is not physically bonded, or mounted to some material which is semi conductive. Reach up and touch the fixture, feel a sharp jolt (discharging the concentration), and the Lamps fire up.
How did everything turn out?
Scott " 35 " Thompson Just Say NO To Green Eggs And Ham!
I guess the exact cause of the problem will remain a mystery because I never got the time to check into this further. We had a lot of loose ends to tie up before the new guy would be in a good position to take over the job. I explained the situation to him and he is aware of the problem. He didn't even seem curious as to what could be causing it though. He said something like, "they're just temporary lights". I suppose he's right, temporary lighting would never cause a problem like the temporary panel plug that went up in flames!