LED with ballasts, spider box 120v keeps tripping powered by a 20KVA Diesel generator

[jhoust]

I'm just an apprentice, be merciful.

Here is the set up:

Power source is a 20KVA Diesel Generator with a total of 5 circuits drawing power from it, all lighting (nothing crazy no power tools running or motors).

~50 ft SO 6/4 feeding the spider box, the spider box has 1 extension 14g 50ft (simple enough)

I have a lighting circuit with 4 can lights and 3 LED light strips (drywall recessed, bathroom)

I ran power off the extension cord to my switch leg and it tripped. I checked each LED light and they work (individually).

I isolated the problem to the LED strips after simplifying my circuit, it's literally just wire MC going to the LED strips.

Here is the problem?:

When I energize just 1 LED the spider box doesn't trip, but when I put two LED strips, it trips immediately.

The LED lights are electronic ballast and they are about 10ft longest (2 are joined to make that length) and the other LED strip is 6 ft with a 1ft extension. These things work when I power them individually.

I don't know how the ballasts were wired, I didn't wire them, but they are suppose to be powered 208-240v? from utility and I am giving them 120v from site generator.


Questions:

Why did my spider box trip? 20amp

Would harmonics play a roll in this problem, why or why not? Generator vs Utility?

What roll do the ballasts play in my problem, are they the cause of my grief? In-rush of current?

What is happening in my circuit? Too much work (amps) and not enough juice (voltage)?


Remarks:

It's not a dead short.

Thank you for looking at my problem. If you want clarification on anything I will respond in the morning, I'm going to bed.

Seriously thank you if you are going to respond to me.

 

[suemarkp]

Is this a regular breaker or a GFCI breaker? The voltage the LEDs want should be on the nameplate for the LED controller. Does it say 120V? I would think you have the voltage right is they work individually.

 

[jhoust]

Is this a regular breaker or a GFCI breaker? The voltage the LEDs want should be on the nameplate for the LED controller. Does it say 120V? I would think you have the voltage right is they work individually.

It's a GFCI breaker on the spider box, I don't know about the nameplate, I would have to return on monday to find out.

 

[suemarkp]

So it is probably the GFCI that is tripping. If you temporarily disconnect the equipment ground at the lights does the breaker still trip? If not, it could be the neutral and ground are touching somewhere, or there is just too much leakage or capacitive coupling and the GFCI is detecting that and killing the circuit. You could also try replacing the GFCI breaker in case it has become overly sensitive, but that probably isn't it.

You could replace the GFCI breaker with a regular one in order to troubleshoot. Use an amp clamp to measure current in the hot, neutral, and ground. Find out where the leakage is. It seems that each lamp leaks some, and when you add enough of them you get to the 5 to 6 mA GFCI trip level and the breaker pops.

I'm not sure how you fix a leakage problem if there is no N-G accidental connection. You may need to discuss with the manufacturer your issue and ask if they have a solution. Usually, fixed lighting doesn't need to be on a GFCI circuit so they may not have checked functionality for that leakage on GFCI circuits.

 

[jhoust]

So it is probably the GFCI that is tripping. If you temporarily disconnect the equipment ground at the lights does the breaker still trip?

I will try this on Monday and give a follow up.

If not, it could be the neutral and ground are touching somewhere, or there is just too much leakage or capacitive coupling and the GFCI is detecting that and killing the circuit. You could also try replacing the GFCI breaker in case it has become overly sensitive, but that probably isn't it.

It's possible that the neutral is leaking with MC cable, but I always use the little red sleeves they give just to avoid this kind of problem. Would be very annoying to me if this is the case. If it's leaking at the LED fixture that would be better. The spider box has seen much in it's time, pretty sure this could also be an issue.

You could replace the GFCI breaker with a regular one in order to troubleshoot. Use an amp clamp to measure current in the hot, neutral, and ground. Find out where the leakage is. It seems that each lamp leaks some, and when you add enough of them you get to the 5 to 6 mA GFCI trip level and the breaker pops.

I will check this out on Monday.

I'm not sure how you fix a leakage problem if there is no N-G accidental connection. You may need to discuss with the manufacturer your issue and ask if they have a solution. Usually, fixed lighting doesn't need to be on a GFCI circuit so they may not have checked functionality for that leakage on GFCI circuits.

I have noticed that when I search the internet about this problem, it always involves a LED light fixture and GFCIs. Lots of electricians out there saying that LEDs will not trip a GFCI and I'm in no position of experience to argue that point, but when I do fix this problem I will report what I found out back here to this thread, even a call to the manufacturer, if it get's to that point.

Thank you for the response!

 

[synchro]

I agree that suemarkp's recommendation to temporarily disconnect the equipment grounding conductor to the lights would be the best thing to try first.

If the breaker still trips after disconnecting the ground wire to the lights, a suggestion would be to put a separate switch on each of two lights just for testing purposes. That way you can turn on the lights one at a time to avoid their inrush currents from occurring at the same time and adding together. So as a test turn on one light, wait a few seconds, and then turn on the other one. If the breaker holds then the inrush current is tripping the breaker, or at least it's pushing it over the edge.

 

[GoldDigger]

jhoust, it may not help in your case, but when you have to use multiple loads, each of which does not excede the ~5ma GFCI limit but added together can cause a trip, there is one way to avoid the tripping.
Instead of powering them all from a single 120V circuit, split them between the two sides of a 120-0-120 MWBC. The result is that the leakage from the two halves will cancel out on the neutral. You would have to use a two pole GFCI to feed the MWBC.
The correct resolution of the problem would be to find LED lighting equipment with a lower leakage between the 120V input and ground, assuming that the GFCI protection is mandatory.

 

[jhoust]

I agree that suemarkp's recommendation to temporarily disconnect the equipment grounding conductor to the lights would be the best thing to try first.

If the breaker still trips after disconnecting the ground wire to the lights, a suggestion would be to put a separate switch on each of two lights just for testing purposes. That way you can turn on the lights one at a time to avoid their inrush currents from occurring at the same time and adding together. So as a test turn on one light, wait a few seconds, and then turn on the other one. If the breaker holds then the inrush current is tripping the breaker, or at least it's pushing it over the edge.

Good Idea thank you for the reply.

 

[jhoust]

jhoust, it may not help in your case, but when you have to use multiple loads, each of which does not excede the ~5ma GFCI limit but added together can cause a trip, there is one way to avoid the tripping.
Instead of powering them all from a single 120V circuit, split them between the two sides of a 120-0-120 MWBC. The result is that the leakage from the two halves will cancel out on the neutral. You would have to use a two pole GFCI to feed the MWBC.
The correct resolution of the problem would be to find LED lighting equipment with a lower leakage between the 120V input and ground, assuming that the GFCI protection is mandatory.

The LED lighting is not suppose to be on a GFCI, but the spider box (it's GFCI), I will probably throw this on a regular breaker first and try it, while reading the current through a clamp, just to make sure everything is peachy. If I'm still getting some trouble I will investigate further after reading my current on all my conductors.

The thing is when I search online for "GFCI tripping with LED lights", I get a lot of "electricians" saying they never have had LEDs trip a GFCI before. Maybe they didn't have electronic ballasts controls on their LEDs, maybe they only had one LED strip on the GFCI, but I'm starting to think that the GFCI is the main problem here and it's not playing well with the inrush current of two LED strips with electronic ballasts. I want to put this to rest, hopefully it's just the GFCI, which isn't going to be part of the circuit when the new construction is finished.

I will be posting what I find out on Monday here after work.

 

[jhoust]

Went into work today and switched out the spider box with a better one. I guess the spider box that I was using once had several days of rain dump on it. Circuit worked immediately. It was maddening, because spider boxes have several GFCIs that are on different circuits and the ENTIRE spider box was bad.


Go figure.

 

[GoldDigger]

Went into work today and switched out the spider box with a better one. I guess the spider box that I was using once had several days of rain dump on it. Circuit worked immediately. It was maddening, because spider boxes have several GFCIs that are on different circuits and the ENTIRE spider box was bad.


Go figure.

Or possibly the GFCI breakers in one box came from a run which tripped on the low end of the allowed range while the ones in the newer box were on the high end of the range.
But it is certainly possible that there was high leakage on all the circuits of the old box because of its history and a small extra current could cause a trip.

 

[ptonsparky]

Went into work today and switched out the spider box with a better one. I guess the spider box that I was using once had several days of rain dump on it. Circuit worked immediately. It was maddening, because spider boxes have several GFCIs that are on different circuits and the ENTIRE spider box was bad.


Go figure.

Thank you for the update.

 

[jhoust]

Or possibly the GFCI breakers in one box came from a run which tripped on the low end of the allowed range while the ones in the newer box were on the high end of the range.
But it is certainly possible that there was high leakage on all the circuits of the old box because of its history and a small extra current could cause a trip.

Absolutely possible as well, the spider box I first had my circuit plugged into had lots of wear on it (I'm not sure how this would affect the device), the other spider box had brand new GFCIs, so they are years apart and different designs (I think). I wanted to put this circuit on a regular breaker anyways, it's nice to know that it doesn't trip on the new GFCIs. I don't have an Amp clamp (just voltage), but I think I will buy one and "see" what the leakage is on these lights, just to nerd out on the other guys. Also, that old spider box was underwater for a couple days too this rainy monsoon season, (I didn't know this until another apprentice told me after troubleshooting) pretty sure that didn't do it any favors.

 

[synchro]

... I don't have an Amp clamp (just voltage), but I think I will buy one and "see" what the leakage is on these lights, just to nerd out on the other guys.

Most clamp meters don't have enough resolution to accurately measure leakage currents of a few milliamps.
If you want to get one with sufficient resolution, I suggest searching online for a "leakage clamp meter" to find something in your desired price range.

 

[jhoust]

Most clamp meters don't have enough resolution to accurately measure leakage currents of a few milliamps.
If you want to get one with sufficient resolution, I suggest searching online for a "leakage clamp meter" to find something in your desired price range.

Thank you for the heads up on that, I didn't know the sensitivity on them, I was just now looking around for one.