Semi-DIY Solar Generator Issue - GFCI outlet tripping

[RocketTier]

Hello, all. I am new to posting, but have been lurking for a while.

Issue: Repeatedly tripping a GFCI outlet once I pass ~600w of power draw on a small(ish) semi-DIY solar gen. Unsure if it is a G-N bond issue, but am more concerned that I have an actual current difference between the neutral and hot legs caused by a short somewhere in my system.

Background: I recently put together a 48v solar generator/backup system, closely following many of Will's videos, as well as many of the discussions spread throughout these forums. I am using an EG4 6500ex, and a single 48v EG4-LL (V1) battery. I have a PV DC-isolator in between the all-in-one and my main PV line, and have wired the AC output to a 20Amp GFCI outlet as I felt uneasy just using a 14ga extension cord. I've been in construction for years and have done plenty of small repairs and DIY electrical work, but I am still definitely a beginner with all of this and am honestly pretty nervous about working with electricity on something that depends on so much of my own work and on connections that I can't visibly see (the ones enclosed in the 6500), so I have been slowly increasing the amount of solar input and AC output as I test the setup. I got up to 900w/~160vOC of PV input, and that has been charging the system great and with no faults showing in the 6500 itself. However, whenever I attempt to draw more than about 600w of power (it varies each time, that has just been the average), the GFCI outlet trips within seconds.

I have the older 6500ex from before the N-G bonding issue was remedied, and to be honest, I still don't fully understand the issue even though I have watched videos on it and have read threads here discussing it. I thought my setup was small and simple enough to not run into problems with it, but I'm hoping it really is related to that and not something more dangerous/concerning.

I'm sure I have used incorrect terminology in this post in my ignorance, so please go easy on me but I welcome and appreciate corrections and guidance and any requests for more details.

EDIT: I may just be stupid for using a GFCI outlet and should just opt for using a 12ga extension cord, but I just don't see why I would be having this issue unless I have a more serious problem that I just don't understand and need to address. But again, ignorance.

 

[TacomaJoe]

I'd try two things. First, try a different load to be sure that it is not tripping the GFCI. Second, I would try lifting the ground wire to the GFCI (should run with just hot and neutral) and try your loads.

GFCI trip on the current imbalance on the output side. They also trip if there is a ground to neutral connection on the output side.

 

[RocketTier]

I'd try two things. First, try a different load to be sure that it is not tripping the GFCI. Second, I would try lifting the ground wire to the GFCI (should run with just hot and neutral) and try your loads.

GFCI trip on the current imbalance on the output side. They also trip if there is a ground to neutral connection on the output side.

Awesome, thanks for the suggestions!

By "lifting" the ground wire, does that include complete disconnection of the ground from both the inverter and the GFCI?
Since my setup isn't technically grounded, is there a reason to have the ground wire in place other than for some kind of fault sensing in the inverter? Meaning, if I lift the ground wire and it solves the problem, do I then have to find the deeper issue so I can still replace the ground eventually running to the GFCI?

 

[TacomaJoe]

I'm not sure how a GFCI handles a floating neutral. I would guess that it doesn't care but I'm a bad guesser sometimes. By lifting the ground wire from the GFCI that would take it out of the equation. I don't know how your inverter handles a neutral/ground bond. A neutral/ground bond may make the GFCI happy.

My thinking is that as you load the inverter, the ground to neutral voltage is changing (because it is floating) and the GFCI trips because of it. Lifting the ground wire, or bonding the neutral and ground together would take any voltage difference issues out of the equation.

Have you tried loading the inverter without the GFCI?

 

[RocketTier]

I'm not sure how a GFCI handles a floating neutral. I would guess that it doesn't care but I'm a bad guesser sometimes. By lifting the ground wire from the GFCI that would take it out of the equation. I don't know how your inverter handles a neutral/ground bond. A neutral/ground bond may make the GFCI happy.

My thinking is that as you load the inverter, the ground to neutral voltage is changing (because it is floating) and the GFCI trips because of it. Lifting the ground wire, or bonding the neutral and ground together would take any voltage difference issues out of the equation.

Have you tried loading the inverter without the GFCI?

I think I'm tracking, now. I think I would have to get the set screw for the 6500ex to activate the neutral/ground bond, so I may be out of luck with that one.

I haven't tried loading it without the GFCI, so I'll probably wire a new non-GFCI outlet and see if I start throwing any faults in the inverter.

 

[TacomaJoe]

Lacking a set screw, you can also jumper across at the GFCI input.

 

[zanydroid]

What is the load connected to the GFCI? Does it internally bond G-N (like it shouldn’t). Does it trip other GFCI in the house?

Basic GFCI measures L-N imbalance. They actually do not need to be grounded to work, and that is a code compliant way to improve safety of ungrounded receptacles.

I agree with disconnecting the ground leading to the load to eliminate one possible source of the imbalance current. But this is just for diagnosis to get more insight into WTF is going on. Even if this makes the GFCI reliable there is still strong evidence that something untoward is happening in the load and you should be careful.

 

[RocketTier]

What is the load connected to the GFCI? Does it internally bond G-N (like it shouldn’t). Does it trip other GFCI in the house?

Basic GFCI measures L-N imbalance. They actually do not need to be grounded to work, and that is a code compliant way to improve safety of ungrounded receptacles.

I agree with disconnecting the ground leading to the load to eliminate one possible source of the imbalance current. But this is just for diagnosis to get more insight into WTF is going on. Even if this makes the GFCI reliable there is still strong evidence that something untoward is happening in the load and you should be careful.

The load is a single window unit AC. I haven't tried other GFCI outlets as it isn't anywhere near them, but I suppose I should try it with a heavy duty extension cord.

I wired in a non-GFCI output and now the loads are running just fine. I haven't thrown any faults in the inverter, and the AC is running strong with no issues. I should have tried lifting the ground before completely disconnecting the GFCI, but I was concerned about doing that before I knew more about it and was more concerned about the load itself first. But again, I should have done that first for more diagnostic knowledge.

I guess now I just need to decide if it is worth rewiring the GFCI back in and figuring out the issue or just continuing with the non-GFCI. I've been obsessively checking every pieces of metal on the setup with my multimeter looking for current anywhere it shouldn't be and haven't been able to find it, so I feel ok so far leaving as is, but I'm still open to any suggestions.

 

[zanydroid]

There are also plug-in GFCI adapters: TRC Southwire 14650013-6 120-Volt, 15-Amp, 1800-Watts, Single Outlet GFCI Adapter, for Indoor Use with Manual Reset, Yellow https://a.co/d/hJuw1kV

You can use this to test with a non GFCI receptacle in the house.

Is it an inverter AC or classic design? Inverter power supplies have ground leakage during regular operation and not all have been redesigned to work with GFCI.

What kind of room/flooring? If it’s a regular room where GFCI is not required by code, I would probably not worry about it, and make sure the power is unplugged before moving the AC. If it is garage, basement, or concrete /dirt then having working GFCI is a good idea.

Another approach is to install GFCI and non GFCI receptacles after your inverter and use GFCI by default.

When no GFCI is in place you want to be extra careful about N-G bond for your inverter output so that line to chassis faults are cleared.

I've been obsessively checking every pieces of metal on the setup with my multimeter looking for current anywhere it shouldn't be and haven't been able to find it, so I feel ok so far leaving as is, but I'm still open to any suggestions.

How are you testing current with a multimeter? I thought it needs to be done in series.

 

[hwy17]

I would at least try a different load first, like a space heater, to figure out if this is just the A/C resulting in this or any appliance.