GFCI Trip on Plug in (spare connector)

[S-2000 Roadster]

My understanding, and someone can correct me if I'm wrong, is that while a circuit breaker detects excess current, a GFCI detects imbalanced current: more current going out than coming back, which means there is leakage to ground. Also, it is my understanding that a GFCI will trip much faster than a circuit breaker.

That's basically correct, but I think it's important (again) to consider the safety ground.

At a minimum, you need two contacts to make a circuit, but you'll note that high-powered electronics generally have three contacts. Technically, one of the two blades on a 120 VAC plug is neutral, which is basically the same as ground. That neutral contact should be carrying 100% of the return current going back to the socket. The safety ground is there to conduct current (hopefully instead of your body) when there is a fault in the high-power device. If the device is working correctly, there should be 0 A flowing through the safety ground contact. The GFCI circuit looks for 0 A, plus a small threshold for slight imperfections, and trips long before the breaker if it finds current on the safety ground.

In other words, whenever there is more current going out than coming back, this imbalance is between the hot and neutral contacts. But, according to the rules of electronics, there must always be a circuit, so the safety ground ends up passing the difference in current. GFCI can detect an unexpected current in the safety ground and/or an imbalance in current between hot and neutral. Note that another possibility for an imbalance occurs when there is another path for ground current, such as through the 'earth' or a body of conductive water. That's an unsafe ground, and would not show up in the safety ground current measurement, but would still cause an imbalance.

As for the Xebra, it has an aftermarket LiFePO4 battery pack and matching charger. It draws 25 amps at 120 volts. The EV Porsche, however, (which was sort of operational for a month before it went in to another shop to be fixed from the original botched conversion job) charges off an RV-style circuit: 240 volts, 50 amps, though my understanding is that it only draws 40 amps. IOW it draws the same as the Roadster when the Roadster is on the UMC. In fact, I had the circuit installed for the Porsche. It was fortuitous that it's perfect for the Roadster as well. The Porsche, like the Xebra, plugs directly into the wall and that connects directly to the charger. But unlike the Roadster, neither the Porsche nor the Xebra can plug into anything but their respective design circuits.

That Xebra LiFePO4 appears to be a 3,000 W charger! I'm surprised that you can even safely plug that in to a standard 120 VAC receptacle, considering that the highest ratings are 12A/15A and 16A/20A. I was under the impression that an 1800 W hair dryer was about the most powerful thing you can plug in at home (maybe 1920 W at the extreme). Where, exactly, do you plug in this Xebra?

By the way, there is a very specific reason why all of these 50 A chargers only draw 40 A. Electrical codes state that normal use should only pull 80% of the rated current. I learned this from a certified electrician at Lowes, as well as from advice here on this forum when I installed my own NEMA 14-50. If you do the quick math, 80% of 50 A is 40 A. In fact, every one of the available current presets on the Tesla Roadster is 80% of a standard breaker. The only exception is that the 90 A Tesla HPC should be able to draw 72 A instead of only 70 A, but the limit is the rating of the connector on the Roadster itself.

Another thing to note is that the plugs and receptacles have current ratings, too. So, that NEMA 15-50 connection is only rated at 50 A maximum. If you boost the breaker and set the Roadster to pull 70 A, then everything will be fine except the plug and socket, which might catch fire! Getting back to your Xebra, I'm surprised that you can plug anything in, since standard extension cords and 120 V sockets are not really rated for 25 A.

 

[TEG]

And something else I recall about GFCI... I think some use 5ma min leak before tripping and others use 20ma... The ones that only need 5ma to trip tend to be overly sensitive and can switch off just from some line noise from things like fluorescent light start-up and such. Not all GFCIs are the same, and not all are the same sensitivity.

 

[daniel]

At a minimum, you need two contacts to make a circuit, <snip> The safety ground is there to conduct current (hopefully instead of your body) when there is a fault in the high-power device. If the device is working correctly, there should be 0 A flowing through the safety ground contact. The GFCI circuit looks for 0 A, plus a small threshold for slight imperfections, and trips long before the breaker if it finds current on the safety ground.

In other words, whenever there is more current going out than coming back, this imbalance is between the hot and neutral contacts. But, according to the rules of electronics, there must always be a circuit, so the safety ground ends up passing the difference in current. GFCI can detect an unexpected current in the safety ground and/or an imbalance in current between hot and neutral. Note that another possibility for an imbalance occurs when there is another path for ground current, such as through the 'earth' or a body of conductive water. That's an unsafe ground, and would not show up in the safety ground current measurement, but would still cause an imbalance.

The GFCI will trip if there is an imbalance between current flowing out and current flowing back in. That includes a situation when you touch a live wire and you become the return to earth. The safety ground (which is important, as you say) protects you if there is a short within the device. But if you touch a frayed wire on the extension cord between the device and the wall, where the fault is in the cord and not the device, or if the device has an exposed live wire and you touch that, it will trip, and it will do it much faster than a circuit breaker, which, being designed to protect the wires, will allow some amount of excess current for some brief time, since the wires can tolerate a little excess for a brief time.

That Xebra LiFePO4 appears to be a 3,000 W charger! I'm surprised that you can even safely plug that in to a standard 120 VAC receptacle, considering that the highest ratings are 12A/15A and 16A/20A. I was under the impression that an 1800 W hair dryer was about the most powerful thing you can plug in at home (maybe 1920 W at the extreme). Where, exactly, do you plug in this Xebra?

The Xebra plugs into a 120 volt 30 amp circuit. I don't remember the NEMA designation of the plug, but it's a twist-lock plug rated for the amperage. An electrician installed it by taking half the dryer circuit. I cannot charge the Xebra while the dryer is running, which has never been a problem, since I only do laundry once every week or ten days, and I just don't charge the Xebra then.

 

[Mycroft]

The GFCI will trip if there is an imbalance between current flowing out and current flowing back in.

Yep. GFCI is no help at all in the event of a short. Don't ask me how I know.:redface:

 

[S-2000 Roadster]

But if you touch a frayed wire on the extension cord between the device and the wall, where the fault is in the cord and not the device, or if the device has an exposed live wire and you touch that, it will trip, and it will do it much faster than a circuit breaker, which, being designed to protect the wires, will allow some amount of excess current for some brief time, since the wires can tolerate a little excess for a brief time.

I'm just nit-picking here, but breakers are designed to protect more than just the wires. They're designed to prevent fires and also to prevent electrocution. Granted, GFCI protects against electrocution in situations where breakers (and the former technology, fuses) would not. I just want to point out that breakers were not specifically designed to favor protecting wires instead of protecting humans. In fact, their primary purpose is to work like a fuse but to add the feature of being resettable rather than burning out and having to be replaced. If you've ever lived in a residence with a fuse box, you'd understand the benefit. But these are still rather simple technologies, and thus GFCI was invented to improve the protection - it's not like GFCI is the first time anybody thought of preventing electrocutions. Unfortunately, GFCI is new technology and it still doesn't work as smoothly as the simpler technology. At least it errs on the side of over-protecting...

Also note that the toleration for brief current overages is part of the nature of the simplicity of the technology. It's not so much a purposeful feature that sacrifices humans in favor of more resilient wires, but rather it's just a coincidental side-effect.

 

[TEG]

Well I still think a plain old fuse or circuit breaker doesn't offer much electrocution protection. It could stop a wire from melting the insulation thus avoid the shock hazard in the first place, but if you come in contact with the wires I doubt the breaker is going to offer you much protection.

Related:
Residual-current device - Wikipedia, the free encyclopedia

...RCDs are designed to prevent electrocution by detecting the leakage current, which can be far smaller (typically 5–30 milliamperes) than the currents needed to operate conventional circuit breakers or fuses (several amperes). RCDs are intended to operate within 25-40 milliseconds, before electric shock can drive the heart into ventricular fibrillation, the most common cause of death through electric shock. In Europe, the commonly used RCDs have trip currents of 10–300 mA...
In the United States, the National Electrical Code requires GFCI devices intended to protect people to interrupt the circuit if the leakage current exceeds a range of 4–6 mA of current (the trip setting is typically 5 mA) within 25 ms. A GFCI device which protects equipment (not people) is allowed to trip as high as 30 mA of current; this is known as an Equipment Protective Device (EPD). RCDs with trip currents as high as 500 mA are sometimes deployed in environments (such as computing centers) where a lower threshold would carry an unacceptable risk of accidental trips. These high-current RCDs serve for equipment and fire protection instead of protection against the risks of electrical shocks...

 

[daniel]

Yes, I've lived in a house with fuses, and yes, the advantage of breakers is obvious. But the fire-protection of a fuse or breaker is a side-effect or a corollary effect of protecting the wires: It's when wires overheat due to excessive current that they can cause fires. Breakers and fuses prevent fires by protecting the wires.

And I'd argue that both fuses and breakers are designed with a delay time, precisely because wires can tolerate excessive current for a short amount of time. We use surge protectors (essentially fast circuit breakers) to protect equipment that cannot tolerate brief spikes. We could put surge protectors in the entry box, but we don't, because the fuses or breakers are intended to protect the wiring, and the wiring can tolerate spikes. Computers and people cannot, so we have surge protectors and GFCIs where needed.

I remember when people used to put a penny behind the fuse, thus bypassing it, when they didn't have a spare fuse handy. It got their power back on, but left them with no protection from the excessive current that caused the fuse to blow in the first place. Circuit breakers have probably prevented many fires because they put an end to imbeciles putting a penny behind the fuse.

 

[Slackjaw]

View attachment 2105
No splicing was done; just detach the GFCI and connect the plug to the cable wires.
The plug was bought at a local Home Depot.
There's also a matching female connector available.

After finding I couldn't plug in to some charge stations on 110v (e.g. ChargePoint) due to the shape/orientation of the connector and socket, I did the same as DaveD, using exactly the connectors linked above. I was actually a bit more agressive and cut the spare connector cable with wire cutters about a foot down. Then I stripped the wires and put the two connectors in, and tested it; full functionality with or without the GFCI dongle.

The only additional work was to file down the little grippy things which pinch the cable at the back of the connector. They needed some quite serious trimming to accommodate the thick yellow cables, but removing them completely would have left it vulnerable to damage if yanked out of the socket or tripped over.

This is only a "last resort" as we've bought the J1772 connector already, but definitely a worthwhile improvement. Thanks as always to the various contributors on this board. It's a great resource.

 

[bonnie]

After having experienced the plug tripping on a roadtrip (resulting in a Plan B charging side trip the next morning), I replaced the plug. Happy to report that the issue is now gone. I did buy a GFCI standalone plug, so if I want to put GFCI on the line, I can.

I wouldn't have been aware of the cause of the tripping, if not for members on this forum. Thanks, all! (And for any new Roadster owner reading this thread, I HIGHLY recommend you make this simple fix to your 110v charging cable - otherwise you could be faced with a GFCI-equipped outdoor plug and be out of luck.)

 

[dsm363]

I did buy a GFCI standalone plug, so if I want to put GFCI on the line, I can.

Do you have a link to this on Home Depot or something similar too? I don't think I'd be able to do what Slackjaw did above but the GFCI standalone plug might work if I change the plug. Thanks.

 

[bonnie]

Do you have a link to this on Home Depot or something similar too? I don't think I'd be able to do what Slackjaw did above but the GFCI standalone plug might work if I change the plug. Thanks.

There are a number of them on Amazon - here's a 15 amp:TRC 90265-6-012 Shockshield Yellow Portable GFCI Plug with Surge Protection

 

[Aspirant]

Thank goodness I found this thread. My 15A charger broke and Tesla said I could buy a new one for $600. Turned out it was just that the cable's GFCI had flaked due to being so frequently plugged into an outlet that was also GFCI. $4 in parts later and I'm back in business. I'll add that the cable's original GFCI housing was water proof, and that anyone who modifies their cable should consider using some silicone gel to seal it up.

 

[W.Petefish]

Or use a NEMA rated plug. (think electrical supply shop, not Home Depot.)

 

[wycolo]

'Turned out': according to whom?? I do not see any 'duelling' going on if one connects a GFI containg plug into a GFI containing socket. As I have been doing for the past month with my 'basic cable' plugged into an outside outlet. Two GFIs in series just means the most sensitive device can kill the circuit first. And the other just remains 'on'. I suspect simple hardware issues here. Always nice to replace well-used hardware. Also, silicone can trap moisture that would otherwise dissipate or evaporate.
--

 

[TEG]

Related threads:
Solutions for double GFCI
The GFCI topic

 

[strider]

Bringing up an old thread here....

I replaced the end of my MC120 with a regular plug per this thread and have charged at lots of regular outlets with no problems. My company recently installed a Chargepoint station and the L1 plug has been going unused so I thought I would have my car sip power during the day and leave the L2 for the Volts and Leafs.

However, when I tried to use my MC120 the Chargepoint station threw a GFCI error and wouldn't charge. Anyone seen this? Do I need to figure out a way to add the GFCI back into the mix?

 

[Doug_G]

No, that would mean the CFCI in the Chargepoint station itself tripped. Maybe it has too low a threshold.

 

[strider]

No, that would mean the CFCI in the Chargepoint station itself tripped. Maybe it has too low a threshold.

Ok, thanks. I'll try another station and see if it happens there. I have heard that a Leaf and PiP have charged using the L1 on this station so I think it's working. May ask to borrow someone's L1-J1772 cord and see if that works.

 

[hcsharp]

Ok, thanks. I'll try another station and see if it happens there. I have heard that a Leaf and PiP have charged using the L1 on this station so I think it's working. May ask to borrow someone's L1-J1772 cord and see if that works.

The Leaf and PiP are much less likely to trip the GFCI than the Roadster, especially if you have a 2.5. There's something about the charging circuit in the PEM that will often draw an unbalanced load. I have never been able to charge off of a standard outdoor 120v GFI outlet despite changing the plug per above. I don't know what the ChargePoint outlet leakage threshold is but a standard 120v GFI outlet is 5ma - generally enough for a Leaf or PiP but not enough for a Roadster which needs 20mA. Sadly for us a lot of the cheaper public EVSEs only allow 6mA before tripping.