EV Charger GFCI Problem

[Jcrist]

I have recently installed a set of 6 electric vehicle chargers and everyone is giving me the same code of GFCI failure. The GFCIs are internal to the chargers and are single phase 208. We installed a new sub panel fed from an existing sub panel. I believe after further investigation of the existing system, I have found that the neutral is ungrounded. The neutral is ungrounded in the transformer, but unsure about the main panel do to a bolt that could be bonded to the back of the panel. I have exhausted all other options that come to mind.

Would an ungrounded neutral effect a 208 GFCI that doesn’t use the neutral?

 

[Little Bill]

An unbonded neutral would not effect the GFCI. However, GFCI breakers require the neutral pigtail to connect to the neutral bar or they won't set/reset. I'm not sure how the GFCI part of the charger works. On some breaker brands, their GFCI breakers won't work on 208V. Not sure if the charger's internal GFCI works on 208V. Do the specs for the chargers say for use on 208V?

 

[ActionDave]

Couple of questions,
What does your meter tell you when you test voltage?

Don't forget the transformal neutral can be bonded at the transformer or the secondary disconnect. Did you look both places?

I don't think I would trust the diagnostics on the charger until I knew the wiring was correct.

 

[wwhitney]

Some EVSEs conduct a "ground check" in which they intentionally put a few ma on the EGC to confirm that the unit has a functioning EGC. If the EGC is nowhere bonded to the grounded conductor, this check will fail.

Cheers, Wayne

 

[kwired]

Some EVSEs conduct a "ground check" in which they intentionally put a few ma on the EGC to confirm that the unit has a functioning EGC. If the EGC is nowhere bonded to the grounded conductor, this check will fail.

Cheers, Wayne

If that is what is happening then they are adding to confusion to some extent calling that a GFCI failure. I'd call it "grounding failure" or something like that.

 

[Jcrist]

I am reading 208 phase to phase, and 120 phase to ground. The EV charger has its own internal GFCI. Of course the manufacturer book doesn’t go into specifics on what the GFCI problem is. We have tried a supplemental ground at the new sub panel, but it did not clear the issue.

i looked at both the transformer and the first disconnect. Transformer does not have a neutral bond. The first means of disconnect looks like it has no bond. But there is one bolt that might be through the back of can. I did not have time that day to check for continuity between the neutral and the ground.

You

 

[retirede]

If the GFCI function is internal to the charger, and that is detecting a fault, then the problem has to be in the unit or possibly downstream given that your supply voltage is correct.

Different grounding connections on the supply side won’t change anything.

 

[Dennis Alwon]

Sounds to me like there is an issue with the chargers. Again does the unit state 208V.

If the gfci in the unit does not use a neutral then I wonder if it is using the equipment grounding conductor as a means of running the electronics for the gfci. I also wonder if it is gfp and not gfci but that doesn't really matter.

So I guess I would make sure you have a good equipment grounding conductor all the way thru. I assume you are sure that the unit doesn't need a neutral.

 

[retirede]

Sounds to me like there is an issue with the chargers. Again does the unit state 208V.

If the gfci in the unit does not use a neutral then I wonder if it is using the equipment grounding conductor as a means of running the electronics for the gfci. I also wonder if it is gfp and not gfci but that doesn't really matter.

So I guess I would make sure you have a good equipment grounding conductor all the way thru. I assume you are sure that the unit doesn't need a neutral.

It’s possible that the GFCI circuitry is designed to operate on 208V, no neutral required.

 

[kwired]

Sounds like maybe GFCI protection that is built into a control board and not a stand alone GFCI such as a dead front GFCI mounted within. Such a design could be like was earlier mentioned and it may regularly or even continuously test for an EGC presence as part of it's design. If so I still think it should call a missing EGC something other than a GFCI failure.