Need Electrical Help regarding GFI tripping

[Infinitereef]

Hi all,
I am having a situation here. Let me describe my setup. I have a 180g In wall tank. I have the electrician installed 8 electrical GFI outlets on 1 circuit on the left side of the tank. I did the same on the right side of the tank. So I have 2 separate 20A circuit with 8 GFI outlet each on each side(Left and Right).

Now, here is the situation happened on my right side of the GFI outlet. I connect my 300W heater to one of the outlet on the right side and everything work fine. Then I connect my 250W metal halide ballast on another Outlet and the light will turned on for 5secs and the GFI will trip. If I extend my wire to connect the Metal Halide to the left side of the GFI outlet, it will work fine. In additional, if the right side of the GFI trip, 5 out of 10 times, the GFI outlet on the left side that supposely on a separate 20A circuit will trip as well?? This is puzzling to me,

1. Why the GFI circuit trip on my Metal Halide??
2. Why sometimes the Left side trip together even though it is on the separate circuit
3. Is it ok that I connect my Light on a non-GFI outlet?? I do have 2 extra non-GFI outlet in the fish room. Is it safety hazzard??

Please help. I don't know what to do??

[The Grim Reefer]

What did the electrician say? Sounds like he didn't do a very good job. If he doesn't make it right I would contact another electrician and a lawyer.

[Philwd]

What kind of ballast are you running? Soemtimes noisy ballasts have been known to trip gfi.

I have had a very similar problem. It was when I plugged a halide into the left side then a second halide on the right the right gfi tripped. I still have not got it sorted out. So all my lights are on the left and heaters etc on the right. Working for now.

Phil

[Infinitereef]

Phil,
Is your left and right side on a separate circuit? I am running a Coralvue ballast.


For now, I think I am just going to have the lighting connected to a non GFI circuit until I figure out the root cause.

[The Grim Reefer]

Any problem with the ballast aside it shouldn't be tripping GFCI's on a different circuit. You can run the ballast on a non GFCI outlet but I would make sure it is just noise and not a shorting issue. It is more important the the equipment touching water be on the GFCI.

[Philwd]

Hi Infinite,
Yes 2 separate 15A circuits at the box. Both are on the right column of the box. I also use a eballast from CV. Hmm.

Phil

[The Grim Reefer]

That just sounds B Zarre!

So I asked a guy that does some electrical work (not an electrician though) and he said to check to see that your neutral lines are grounded, otherwise plug the offending item into power strip with RFM and MFM filters on it and see if that solves the problem.

Good luck.

[Infinitereef]

Philwd,
Have you figure out your problem. I have not.

This is what I did. I run at extension from a non-GFI outlet and connect my MH light on it. What strange is that it is still tripping my right GFI circuit!!!

Grim,
Where to get the RFM/MFM filter power strip??

[The Grim Reefer]

Radio shack should have them, Best Buy and so on too. Just look at the box. They should list the filters employed on the strip. Should cost at least 20 bux for a decent one.

[Infinitereef]

Grim : Thx. Will try at best buy.

[Keiths_Reefs]

Originally posted by Infinitereef
. I run at extension from a non-GFI outlet and connect my MH light on it. What strange is that it is still tripping my right GFI circuit!!!

Where in the circuit was the GFI installed? It has been a common practice to install once GFI receptical and then daisy chain other non-GFI recepticals to that one which in effect makes the non-GFI recepticals, GFI recepticals.

Was this all new wiring or did he just replace existing recepticals?

[artis]

I don't know much about reefing, but I do know a small amount about wiring. This sounds strange to say the least.

First, I'm wondering if you have a defective GFCI. I had a pump that was tripping one when I unplugged the pump (Yeah, that's what I thought). A replacement socket fixed that quickly enough.

I'm also wondering, as others have mentioned, if your electrician has them in daisy chain, or at least chained with other sockets. I use this same arrangement. It works well, but surely would be confusing if one didn't know which was the lead device.

Another thought that comes to mind is the wiring to the line and load sides may have gotten crossed or backwards. Its possible to do this and have it not completely obvious if there is a branching circuit, even with wiring fault detection. Unless you are comfortable, have your electrician check things for proper connections.

Of course, if you open anything, turn things off at the main breaker panel in your home. Reefing is exciting enough without it becoming electrifying. Good luck and be safe.

[aquaman222]

I had problems with the 15 amp GFIs tripping and switched to 20 amps and never another problem.

[Infinitereef]

This is new wiring.

The right GFI circuit has 4 outlets. join together The electrician installed the 1st outlet to a GFCI and then diasy chain to the next 3.

I don't understand the line and load line got backwards. If so, is it easy to fix it?

Should I go ahead and change to another GFCI assuming it is bad?

[Infinitereef]

Actually what I don't understand is that how can running the light on a different circuit trip the GFI??

[artis]

Its so risky to venture a guess without physically examing the wiring. This really sounds like a cross wiring or common ground problem, unless the two circuits are in fact daisy chained.

If one examines a GFCI "wall" plug module, you'll find a number of holes, likely near some lateral screws on the main body. These holes/screws should be labled "line" and "load". The line side is the side that is electrically closest to the breaker box. The load side will be used to connect wires to the "downstream" or daisy chained outlets. I should also mention that the outlets themselves are on the "load" side. This seems confusing at first, but recall the purpose of the GFCI is to isolate wiring if a ground fault is encountered. So essentially, the device "disconnects" the line side (hot) from the load side during a fault.

From a wiring stand point, in most cases, if you wire it up wrong it simply doesn't work. That's a good thing, actually. However, with two or more leads to the "load" side, there is a means to cross line and load where the circuit partially functions. This is bad because there is no protection and the GFCI may trip unpredictably. It can be difficult to determine line from load wiring without a volt-ohm meter.

I always hestitate to recommend someone replace an electrical part due to the safety issues that can arise. If you are extremely comfortable with wiring, have a volt-ohm meter (I often use 2 as a precaution) and proper tools then replacement of the device as a trouble shooting matter might be a good idea. If one is not comfortable with that, then its a job for a licensed electrician. In this case, since there may be a problem with the ground at some point in the circuit, that might be advisable in any case. I'll admit to being over cautious.

[Mr. Ugly]

Originally posted by Infinitereef
Actually what I don't understand is that how can running the light on a different circuit trip the GFI??

Originally posted by Philwd
Soemtimes noisy ballasts have been known to trip gfi.

Yep, it's happening with my setup right now too. I even put the e-ballast on a different circuit and had 2 gfci's trip elsewhere in the house. Once the lights are on, I can reset the gfci's and everything stays on.

We have labs at work that have gfci's nuisance trip with electrical noise.

There are some recent posts in the lighting and equipment forum on the gfci topic. Seems like power noise filters can help.

[artis]

I've never seen line noise cause this and have even run large routers on GFCI circuits, but I surely haven't seen everything either. I suppose its possible if the noise involves the ground.

I'd eliminate the potential for safety related causes first, then investigate that noise issue. If you find that's the problem, please post the type filter and such you used to resolve it. I'd like to install some myself so as not to have the GFCI trip while away from home.

[Infinitereef]

Hey guys, thanks for the inputs. Sorry to say this but i felt better that someone else also experience the same problem.

Can you point me to the thread on the lighting forum. I got the Belkin Surge protector power strip with 40dB noise filtering capability and still does not help.

[NeilPearson]

I have 16 T5 bulbs that also trip my GFCI circuits that I got installed. And I have turned them on on one circuit and they have tripped the other circuit... without tripping the circuit they were on. I had a third unprotected outlet beside the outlets I put in so I tried them there with a home depot powerbar type of GFCI... The lights tripped that one too. So now I am running them on the unprotected outlet without a GFCI.

[Mr. Ugly]

Here's part of a discussion I had with one of our electronic techs at work:


> I haven't done a whole lot of testing, but so far it sounds like line noise from the ballast
> circuits fooling the GFCI'S. Could this be like the GFCI problems with SCR(?) stuff in the labs?

Since the GFIs use transistor logic, they can sense pulses that we don't
know about. Consider the fact that a 60 cycle wave is 16 milliseconds
per cycle. An SCR shut off can take as little as .5 microseconds. The
wire in series with that SCR now has a transition of maybe 2 amps over
.5 microseconds. That 'antenna' can throw that pulse an inch or so to an
adjacent wire, inducing a false current never seen by the equipment but
yet seen by the GFI.

We had such problems in one of our labs. The rated current of the equipment
was 45 amps. We put in a 700A box w/ 60 amp fuses. With the SCR pulsing
on and off, the local oscillations in the mega/giga cycle area in the
line reached over 100 amps, but never seen by anything but the fuse box.
The fuse box was seeing about 45 amps, 60 cycle, and another 50 amps of
megacycle current, induced by the SCRs in the laser.

Similarly, nearby lightning strikes can trip GFCI's through induced currents.

[Mr. Ugly]

Originally posted by artis
I've never seen line noise cause this and have even run large routers on GFCI circuits, but I surely haven't seen everything either. It wouldn't be from typical low frequency line noise like from router or tablesaw motors. More likely from high speed switching in high voltage/current electronics. Also some GFCI are more sensitive than others. Some industrial GFCI have settable trip levels to accomodate this kind of problem.

[artis]

Did some chatting about this at work as well. Unfortunately the only thing we are coming up with that is remotely useful was a question.

Why is this not more widespread if its attributable to high frequency line noise? We don't have that many ballast types in use by the reef community and there surely are not that many manufacturers of GFCI components. Could be a combination of components with specific individual behavior, I suppose.

Would like to see others chime in with thoughts on this. Has me a bit worried about having MH on timers if we can expect to see GFCI trips due to cycling the lights.

[Infinitereef]

Originally posted by NeilPearson
I have 16 T5 bulbs that also trip my GFCI circuits that I got installed. And I have turned them on on one circuit and they have tripped the other circuit... without tripping the circuit they were on.

it trip other circuit and not the one on it!! Wow, this is another bizzare situation.

I try my light on another non-GFI circuit in another room, and it trip my same right GFI circuits in my fish room :confused: :mad:

[Mr. Ugly]

I'd been using several Stanley contractor grade GFCI's and the Stanley heavy duty 8 outlet Prostrips on my various tanks.

http://catalog.belkin.com/IWCatProductPage.process?Merchant_Id=&Section_Id=1719&pcount=&Product_Id=102325&Section.Section_Path=%2F%2FRoot%2FStanleyP%2E%2E%2Eessories%2FGFCI%2FContractorGradeGFCI%2F

http://catalog.belkin.com/IWCatProductPage.process?Merchant_Id=&Section_Id=1709&pcount=&Product_Id=104062&Section.Section_Path=%2F%2FRoot%2FStanleyP%2E%2E%2Eessories%2FPowerStrips%2FHeavyDut%2E%2E%2EhopPowe r%2F

I picked up a canopy with 2x 250MH and VHO actinics all on e-ballasts. My MH e-ballasts would trip the GFCI's on power up. It didn't matter if I plugged the ballasts into the GFCI's or if I plugged them into entirely different circuits. The GFCI's would trip almost every time with even just one of the ballasts. I had this problem with 3 different sets of these GFCI/powerstrips.

I'd read of others with similar problems... even regarding the separate circuits issue. I was very suspicious of an electronic noise issue, having run across similar problems in laboratories at my workplace.

Knowing that excessive cable length can contribute to GFCI problems, I shortened the cords on the strips, the GFCI's, and the ballast outputs the MH. Still, the GFCI's tripped.

I knew that some GFCI's for industrial enviroments have higher trip settings to protect against nuisance trips from line noise. I'd also read that some GFCI's are more sensitive than others.

I decided to try a different GFCI, and picked up a Leviton 20A Smartlock from Home Depot. I wired it up in a 2x4 box with a cord and plug.

I was happily surprised to find that I could plug all of my tank equipment including the lighting into the Leviton without tripping. I even could light both MH at the same time without problem. Interestingly enough, the Stanley GFCI on a different circuit continued to trip while the Leviton stayed on. (Yes, I wired and tested the Leviton correctly with power to the line side.)

So anyway, thought I'd pass along the info regarding noisy ballasts and/or overly sensitive GFCI's.

[Acroholic]

The only way to prevent "Phantom Tripping" on GFCI outlets is to isolate each one. It really has nothing to do with noise or length of cables. Each GFCI must be isolated by running a seperate line for each receptacle. Otherwise if they are run in series, then you will encounter "Phantom Tripping" every time. The best way to do this is to run a subpanel with its own breaker to the tank area. This way you can then wire isolated GFCI receptacles directly from the subpanel. You can also use 12/3 cable instead of 12/2 and get two isolated receptacles from each cable run.

[Mr. Ugly]

Mine were not in series.

[Acroholic]

There is a seperate line from each GFCI receptacle back to the main circuit breaker panel? What guage wires are run?

[Mr. Ugly]

These were Stanley GFCI pigtails. They were on different circuits. They would be in series only if you plugged the pigtails into each other. Even if they were plugged into the same circuit, they would be in parallel.

[Acroholic]

So basically they ARE using common wires....which is causing the "phantom Tripping". The only way to truly have an isolated GFCI receptacle is not to use any common wires whatsoever....

[Mr. Ugly]

separate circuits

separate hots

separate neutrals

separate grounds

gfci's not in series

gfci's not in parallel

ballast on load side of Stanley gfci trips

ballast on load side of Leviton doesn't trip

ballast in parallel with Stanley gfci causes trip

ballast in parallel with Leviton gfci doesn't cause trip

ballast on separate circuit from panel trips Stanley gfci on other breaker

ballast on separate circuit from panel doesn't trip Leviton gfci on other breaker

[Acroholic]

uhhh...okay?

[Infinitereef]

Okay, I will try to change mine out as well to see whether it fix my problem. Thanks for sharing. Seem like light at the end of tunnel for me.

[Mr. Ugly]

Originally posted by Infinitereef
This is new wiring.

The right GFI circuit has 4 outlets. join together The electrician installed the 1st outlet to a GFCI and then diasy chain to the next 3.

I don't understand the line and load line got backwards. If so, is it easy to fix it?

Should I go ahead and change to another GFCI assuming it is bad?
As CapeCoral wrote, make sure you do not have multiple GFCI in series, or daisy chained in the same circuit or you WILL have probs.

You can daisy chain non GFCI receptacles after ONE GFCI. Or you can run multiple GFCI in parallel.

You can have probs from defective/sensitive GFCI or incorrect wiring. As Aquaman wrote, he fixed his by going to a different(20A) GFCI. Mine is ok now that I switched to 20A Leviton smartlocks, which seem to be less sensitive. I don't think my previous ones were defective. I had 3, and they were all new.

Good luck! Please let us know how it goes.

[Mr. Ugly]

Just a followup on the Stanley pigtail GFCI's:

http://catalog.belkin.com/IWCatProductPage.process?Merchant_Id=&Section_Id=1719&pcount=&Product_Id=102325&Section.Section_Path=%2F%2FRoot%2FStanleyP%2E%2E%2Eessories%2FGFCI%2FContractorGradeGFCI%2F#

I RMA'd the devices and received the replacement ones yesterday. They have the same exact trip problem as the previous ones. Looks like they just pulled 3 new ones off the shelf and didn't bother to figure out what was wrong.