Do It Yourself: 240v from two 120v sources

[QueenBee]

EVSEs are required by the J1772 specification to include their own GFI circuitry.

Are you sure this is true? Rich Rudman claims it is not a requirement and that if he doesn't have to add it to get his Manzanita Micro P3 EVSE NRTL listed he will not do so.

One of the reasons I like Tony's/Quick Charge Power's JLONG (J1772 extension cord) is that it helps provide more of your length in GFCI. So for example if you use a Quick220 in the house (because most garage receptacles should have GFCI) you may normally need an extension cord to get the EVSE close enough to the car but if instead you keep the EVSE in the house as well and use the JLONG to get to the car the entire length outside the house is GFCI protected by the EVSE.

[nsayer]

EVSEs are required by the J1772 specification to include their own GFI circuitry.

Are you sure this is true? Rich Rudman claims it is not a requirement and that if he doesn't have to add it to get his Manzanita Micro P3 EVSE NRTL listed he will not do so.

Well, I don't know what certifying organizations (like UL) will care about, but in looking at the October 2012 revision of the J1772 PDF I see a GFI in the diagrams on page 26 and 32 (which depict a L2 and a L3 EVSE/EV block diagram).

Additionally, on page 57 there's a note that talks about an EVSE experiencing an error condition "such as a ground fault."

The document I have doesn't quite explicitly state a GFI requirement, but it's heavily implied.

That's certainly enough for me to suggest that an EVSE that does not include a GFI is not fully compliant with the specification. And building an EVSE without a GFI would, IMHO, be a stunningly bad idea.

[QueenBee]

EVSEs are required by the J1772 specification to include their own GFI circuitry.

Are you sure this is true? Rich Rudman claims it is not a requirement and that if he doesn't have to add it to get his Manzanita Micro P3 EVSE NRTL listed he will not do so.

Well, I don't know what certifying organizations (like UL) will care about, but in looking at the October 2012 revision of the J1772 PDF I see a GFI in the diagrams on page 26 and 32 (which depict a L2 and a L3 EVSE/EV block diagram).

Additionally, on page 57 there's a note that talks about an EVSE experiencing an error condition "such as a ground fault."

The document I have doesn't quite explicitly state a GFI requirement, but it's heavily implied.

That's certainly enough for me to suggest that an EVSE that does not include a GFI is not fully compliant with the specification. And building an EVSE without a GFI would, IMHO, be a stunningly bad idea.

Could not agree with you more. I tried really hard to convince him of that and that he should be warning his customers so that if they are concerned with safety they can at least install a GFCI breaker.

[chris1howell]

GFCI is required by NEC 625.22 and UL 2231. It is mentioned in SAE J1772. Not including it in a product sold to the public is irresponsible.

Are you sure this is true? Rich Rudman claims it is not a requirement and that if he doesn't have to add it to get his Manzanita Micro P3 EVSE NRTL listed he will not do so.

[QueenBee]

GFCI is required by NEC 625.22 and UL 2231. It is mentioned in SAE J1772. Not including it in a product sold to the public is irresponsible.

Are you sure this is true? Rich Rudman claims it is not a requirement and that if he doesn't have to add it to get his Manzanita Micro P3 EVSE NRTL listed he will not do so.

Would NEC be happy with a GFCI circuit breaker or does it require the EVSE it self to also provide it? I have a feeling he'll never get it NRTL listed but glad to see that they'll require it.

I tried very hard to convince him to warn people that he was selling what I consider an unsafe product and to add GFCI but I can understand why he doesn't want to advertise that his product is unsafe and to make it safe would cost significantly more (240 volt GFCI breaker) than it already does compared to his competitors safe products. As to why he doesn't want to add GFCI I have no idea...

[nsayer]

Adding a GFI circuit to an EVSE doesn't cost a lot. It's a single chip, a handful of discrete components and a CT coil. It's like an auto maker complaining that it's too expensive to add brakes to an EV. Sure, you could make a car that stops with regenerative braking entirely, but there's just no way anyone would call the result roadworthy.

The NEC isn't concerned with anything that plugs in. That's why UL gets involved. The NEC talks about EVSEs because some of them are hard wired. In principle a hard wired EVSE with a GFI breaker would be ok.

I don't know what UL's stand is. I'm not sure you could pass by just documenting a requirement for a GFI breaker on the socket.

[chris1howell]

Nope it is required to be in the EVSE for both L1 and L2 within 18" of the plug (if corded). Adding GFCI is not very costly, the OpenEVSE design is about $1 for components and $12 for the coil...

GFCI is required by NEC 625.22 and UL 2231. It is mentioned in SAE J1772. Not including it in a product sold to the public is irresponsible.

Are you sure this is true? Rich Rudman claims it is not a requirement and that if he doesn't have to add it to get his Manzanita Micro P3 EVSE NRTL listed he will not do so.

Would NEC be happy with a GFCI circuit breaker or does it require the EVSE it self to also provide it? I have a feeling he'll never get it NRTL listed but glad to see that they'll require it.

I tried very hard to convince him to warn people that he was selling what I consider an unsafe product and to add GFCI but I can understand why he doesn't want to advertise that his product is unsafe and to make it safe would cost significantly more (240 volt GFCI breaker) than it already does compared to his competitors safe products. As to why he doesn't want to add GFCI I have no idea...

[nsayer]

Nope it is required to be in the EVSE for both L1 and L2 within 18" of the plug (if corded).

I always wondered why the cords were so short. I've been putting, like, 6 foot SOOW cables on mine, but I guess I'm not submitting mine to UL any time soon.

[MikeD]

In "Feature Comparison" advertising I notice Siemens touts their VersiCharge EVSE (along with Schneider EVlink) as having a "UL Listed 5 mA CCID" feature, but lists GE Wattstation, Leviton Evr-Green 160, and Aerovironment EVSE-RS EVSE models as NOT having that feature.

It may be that those three models have a CCID (Charging Circuit Interrupting Device) feature, but at a higher threshold than the "Class A" 5 ma +/- 1 ma level. If so, a higher threshold CCID feature would seem to me to not be quite as safe.

Rather than get lost in discussion about the above or other issues, I want to applaud what nsayer had to say about the important safety value a GFCI protected outlet/circuit provides for plug-in EVSEs in damp conditions.

[nsayer]

In "Feature Comparison" advertising I notice Siemens touts their VersiCharge EVSE (along with Schneider EVlink) as having a "UL Listed 5 mA CCID" feature, but lists GE Wattstation, Leviton Evr-Green 160, and Aerovironment EVSE-RS EVSE models as NOT having that feature.

It may be that those three models have a CCID (Charging Circuit Interrupting Device) feature, but at a higher threshold than the "Class A" 5 ma +/- 1 ma level. If so, a higher threshold CCID feature would seem to me to not be quite as safe.

Rather than get lost in discussion about the above or other issues, I want to applaud what nsayer had to say about the important safety value a GFCI protected outlet/circuit provides for plug-in EVSEs in damp conditions.

Thanks.

I am not sure what the controlling document for EVSE GFI sensitivity is. Chris has stated, and designed OpenEVSE to, a 20 mA standard. I built one for 10 mA, but my wife's Volt trips at that rate just often enough to be annoying. I've personally decided to split the difference and design going forward for something like 15 mA or so.

My assumption is that the switching supply startup in the Volt has an inrush characteristic that momentarily causes an imbalance of just over 10 mA. I'm not such an expert at analog and power electronics design analysis to understand how that might happen. But it's suggestive to me that if Siemens is bragging about 5 mA sensitivity, they're going to piss off a lot of Volt owners unless they're being sophisticated about intensity vs duration of the trip (OpenEVSE and the Hydra don't - any excessive excursion causes a fault response).