60A Breaker -> 4 gage or 6 gage wire?

[eprosenx]

Yeah, I hear ya. GFCI breakers don't even need a ground, but in the case of EVSE with self test of the ground circuit, there could be nuisance faults due to questionable conduit connections which would work fine for phase faults.

@Spartan86, as an FYI, the manual states:
Branch Circuit Conductors and Ground Wire
If installing for less than maximum power, refer to local electrical code to select correct conductors and ground wire size that are suitable for the chosen circuit breaker.
If installing for maximum power, use minimum 6 AWG, 90° C-rated copper wire for conductors.

This can be read to say the ground must also be 6 AWG, if you are setting it up at a 60 Amp feed. However, the heading calls out ground wire vs circuit conductor, so the 6 AWG may only relate to the phase conductors.

On the GFCI breaker bit: GFCI breakers rely on the neutral and ground bond being in place somewhere in the electrical system so that if something faults to ground some current will flow on that path, imbalancing the GFCI circuit and causing it to trip. So yeah, as you point out, the Wall Connector needs a solid ground connection so that it can test that the neutral/ground bond exists. Also, if you are running in metal conduit, having the conduit solidly grounded is critical since if a wire in the conduit melts you want it to fault to the conduit and immediately trip the overcurrent circuit breaker (which is not a GFCI in the case of a Wall Connector).

On the conductor sizing note: Yeah, I think that might be slightly confusing. There is no way Tesla is asking for a full 6 AWG ground. Typically above 10 gauge wiring the ground wires are allowed to be much smaller than the current carrying ones. In the NEC 10 gauge is good for up to a 60a circuit.

I also am totally confused by the requirement for 90C rated wire. This must be a quote from the latest generation Wall Connector manual? (I don't remember it being in the last gen one) Pretty much everything is 90c rated these days (Romex wire is even though code restricts you to the 60c rating, and generally all THHN is). But we are basically always limited to the 75c rating of the wire when connecting to residential circuit breakers which are pretty much universally limited to 75c. So I don't think the Tesla Wall Connector is saying its terminals are 90c rated, but perhaps instead they are saying that they just want to add an additional margin of safety by requesting 90c rated wire regardless?

 

[qdeathstar]

You are penny wise and dollar foolish.

How so? So you have information to back up your uniformed claim I’m assuming?

 

[mongo]

On the GFCI breaker bit: GFCI breakers rely on the neutral and ground bond being in place somewhere in the electrical system so that if something faults to ground some current will flow on that path, imbalancing the GFCI circuit and causing it to trip. So yeah, as you point out, the Wall Connector needs a solid ground connection so that it can test that the neutral/ground bond exists. Also, if you are running in metal conduit, having the conduit solidly grounded is critical since if a wire in the conduit melts you want it to fault to the conduit and immediately trip the overcurrent circuit breaker (which is not a GFCI in the case of a Wall Connector).

On the other hand, if the entire power system is not earth bonded, you can't have a ground fault. Which can be good since you can't make a circuit through your body from one phase, but bad in that if you did get energized between the phases, there is no chance of some leakage to ground tripping the circuit.

On the conductor sizing note: Yeah, I think that might be slightly confusing. There is no way Tesla is asking for a full 6 AWG ground. Typically above 10 gauge wiring the ground wires are allowed to be much smaller than the current carrying ones. In the NEC 10 gauge is good for up to a 60a circuit.

I don't think there are calling for 6 AWG ground either. However, I have dealt with full size ground requirements on hot tubs.

I also am totally confused by the requirement for 90C rated wire. This must be a quote from the latest generation Wall Connector manual? (I don't remember it being in the last gen one) Pretty much everything is 90c rated these days (Romex wire is even though code restricts you to the 60c rating, and generally all THHN is). But we are basically always limited to the 75c rating of the wire when connecting to residential circuit breakers which are pretty much universally limited to 75c. So I don't think the Tesla Wall Connector is saying its terminals are 90c rated, but perhaps instead they are saying that they just want to add an additional margin of safety by requesting 90c rated wire regardless?

Yeah, that is 3rd gen wording. Maybe CYA, or their terminals really are 90C tested (hot garage, full power).

 

[gfunkdave]

On the GFCI breaker bit: GFCI breakers rely on the neutral and ground bond being in place somewhere in the electrical system so that if something faults to ground some current will flow on that path, imbalancing the GFCI circuit and causing it to trip. So yeah, as you point out, the Wall Connector needs a solid ground connection so that it can test that the neutral/ground bond exists. Also, if you are running in metal conduit, having the conduit solidly grounded is critical since if a wire in the conduit melts you want it to fault to the conduit and immediately trip the overcurrent circuit breaker (which is not a GFCI in the case of a Wall Connector).

Not sure I follow here. The GFCI breaker just compares electrons going out to the circuit with electrons being returned through the neutral. If they aren't nearly identical it trips. It doesn't care if the neutral and ground are bonded.

Using the EMT as ground is the standard practice in the Chicago area, where EMT is required in all builds (commercial and residential) and Romex is verboten.

 

[qdeathstar]

I’m not sure about not being able to have a ground fault on systems not bonded to earth... a microwave transforms outputs are not bonded to earth but you can still get electricuted to death by conning between the output and ground (current leakage to ground AKA a ground fault).


The gfci won’t trip in that kind of fault but it can still occur.

 

[eprosenx]

On the other hand, if the entire power system is not earth bonded, you can't have a ground fault. Which can be good since you can't make a circuit through your body from one phase, but bad in that if you did get energized between the phases, there is no chance of some leakage to ground tripping the circuit.

Hah, well yes, if it was not earth bonded then you can't have a ground fault (with just a single fault), but I think we have pretty much universally decided that bonding one leg of electrical systems to earth is typically a good idea (even if you don't actually run ground wires but instead use GFCI / Residual Current Detectors on all circuits instead like they do in some countries).

There are very special situations (like steel factories) that have modified grounding setups (corner grounded delta, ungrounded, etc...) but those require special supervision and training. Or "high resistance ground" systems as well (they are intended to allow you to continue operations even with a fault so you can schedule the shutdown).

Do note though that Tesla clearly requires all EV charging to happen on solidly grounded systems (I know of one instance in a commercial setting where they had an odd grounding setup and hence could not charge).

Yeah, that is 3rd gen wording. Maybe CYA, or their terminals really are 90C tested (hot garage, full power).

Yeah, probably CYA, but the callout on the terminals is a good one. I do seem to remember that Tesla said their terminals are rated to some really hot temp on the previous gen one. That could certainly help as you describe if the Wall Connector was in a really hot location as you describe. I am sure that is why they ask for it.

Not sure I follow here. The GFCI breaker just compares electrons going out to the circuit with electrons being returned through the neutral. If they aren't nearly identical it trips. It doesn't care if the neutral and ground are bonded.

So in the case of a Wall Connector, the breaker is not GFCI, the GFCI circuit is in the Wall Connector itself.

So if neutral and ground are not tied together *somewhere* in the system then if you come in contact with a hot wire and also ground (or anything else grounded) there will be no potential and current won't flow and hence it won't trip the GFCI. It also won't kill you, UNTIL there is a second fault. At which point, the GFCI would still not trip as long as that second fault is also downstream of the the GFCI since the current flow would still be balanced (albeit some of it flowing through your body).

The moral of the story is that it is super critical the neutral and ground are bonded somewhere in the system. ;-)

Using the EMT as ground is the standard practice in the Chicago area, where EMT is required in all builds (commercial and residential) and Romex is verboten.

Yeah, Chicago is so weird... Lol. Not a ton uses EMT around here (unless exposed runs in commercial). Residential is all 100% romex, and commercial is MC cable wherever they can get away with it in walls (so they don't have to bend EMT).

I honestly don't know how often folks around here (when they do use EMT) then also use the conduit as the ground. Typically if folks do EMT I have always seen them run a ground (but again, as you point out, I don't think it is code required).

I’m not sure about not being able to have a ground fault on systems not bonded to earth... a microwave transforms outputs are not bonded to earth but you can still get electricuted to death by conning between the output and ground (current leakage to ground AKA a ground fault).

The gfci won’t trip in that kind of fault but it can still occur.

Uhh, to get electrocuted you have to complete a circuit. If you have a derived source (say an isolation transformer) that has no leg on the secondary side grounded then if a single fault occurs to ground, nothing will happen and nobody will get shocked since no circuit has been completed. You basically have just converted one of the legs into an earthed leg.

Note that I am not sure if a microwave transformer is an isolation transformer or not... It could be an autotransformer...

 

[Spartan86]

Can someone help me identify this wire? It's 6 ga... is it the NM-B?

View attachment 520115

That certainly looks like 6/3 NM B

https://www.lowes.com/pd/Southwire-...Non-Metallic-Jacket-Wire-By-the-Foot/50101582

 

[qdeathstar]

not bonding the neutral to ground wouldn’t result in an isolation transformer... if it were that easy no systems would be bonded to earth...

 

[Bob Denny]

Agreed, this is a good way to check, without pulling the cover off your main panel (which, if you don't know what you're doing, is very risky)

I did this and, within the limits of my VOM, I measure right at 240 at the panel, with the vehicle charging at 48 A, so the losses in the transformer and primary to the panel are maybe a volt or two. After it charges for about 1 hour, the voltage read on the Tesla Charging UI is 236. with 240 showing on the VOM at the dryer plug. So indeed there is a 4 volt drop from the panel to the car (and that includes the 18' cable from the TWC to the car). Just under 200 watts lost in heat.

I'm starting to come around on this, though. The cable from the TWC to the car also gets warm. I'm guessing it is also #6. The temperature of the conduit and the cable are similar. The warmth makes me really nervous, but then what's good for the cable is good for the conduit?

 

[mongo]

I did this and, within the limits of my VOM, I measure right at 240 at the panel, with the vehicle charging at 48 A, so the losses in the transformer and primary to the panel are maybe a volt or two. After it charges for about 1 hour, the voltage read on the Tesla Charging UI is 236. with 240 showing on the VOM at the dryer plug. So indeed there is a 4 volt drop from the panel to the car (and that includes the 18' cable from the TWC to the car). Just under 200 watts lost in heat.

I'm starting to come around on this, though. The cable from the TWC to the car also gets warm. I'm guessing it is also #6. The temperature of the conduit and the cable are similar. The warmth makes me really nervous, but then what's good for the cable is good for the conduit?

Warm cables or conduits are not a problem.

 

[qdeathstar]

I did this and, within the limits of my VOM, I measure right at 240 at the panel, with the vehicle charging at 48 A, so the losses in the transformer and primary to the panel are maybe a volt or two. After it charges for about 1 hour, the voltage read on the Tesla Charging UI is 236. with 240 showing on the VOM at the dryer plug. So indeed there is a 4 volt drop from the panel to the car (and that includes the 18' cable from the TWC to the car). Just under 200 watts lost in heat.

I'm starting to come around on this, though. The cable from the TWC to the car also gets warm. I'm guessing it is also #6. The temperature of the conduit and the cable are similar. The warmth makes me really nervous, but then what's good for the cable is good for the conduit?

if you are worried about it you should have a liscensed electrician look at it. You are probably OK, but who knows..

 

[Bob Denny]

Warm cables or conduits are not a problem.

Yes, the licensed electrician that installed my TWC said that "warm" is not a problem. I really thank all of you for the info and for confirming that he did it right in the first place. 25' run of #6 THHN in 3/4" conduit... I'm good.

 

[eprosenx]

I did this and, within the limits of my VOM, I measure right at 240 at the panel, with the vehicle charging at 48 A, so the losses in the transformer and primary to the panel are maybe a volt or two. After it charges for about 1 hour, the voltage read on the Tesla Charging UI is 236. with 240 showing on the VOM at the dryer plug. So indeed there is a 4 volt drop from the panel to the car (and that includes the 18' cable from the TWC to the car). Just under 200 watts lost in heat.

I'm starting to come around on this, though. The cable from the TWC to the car also gets warm. I'm guessing it is also #6. The temperature of the conduit and the cable are similar. The warmth makes me really nervous, but then what's good for the cable is good for the conduit?

Which wall connector do you have? Is it the new 60a white WiFi one? That is interesting since it has a lower gauge wire than the old one (which was good to 100a), so yeah, I would expect more loss on that.

Warm is totally fine. Hot is not. Sounds like you are just fine!

 

[qdeathstar]

4 volts at 240v is almost nothing, could even be measurement error in your meter depending on how crappy your meter is...

 

[Bob Denny]

@qdeathstar My VOM is pretty good, with a mirror behind the needle and the scale makes it possible to see 1 volt. And the 6 volts (minus the 2 volts on the VOM) is at the car on the Charging display. Anyway, I'm staying with the #6 in conduit.

 

[Electric Steve]

For what it's worth, the manual on the Gen 3 Wall Charger says 6 AWG, 90° C-rated copper wire.

That's not to say that switching to 4 gauge is a bad idea. The price difference appears to be minimal for short runs, and I'd rather have my wires be stone cold.

 

[dcgizmo]

I'm jumping on this thread, but I have to wonder why everyone is talking about, for the NM-B / Romex options, 4/3 or 6/3 instead of 4/2 or 6/2? You only need 2 power wires plus the ground wire. If you use 4/3, you are wasting one of the conductors and making the wire stiffer and harder to work with than necessary.

 

[mongo]

I'm jumping on this thread, but I have to wonder why everyone is talking about, for the NM-B / Romex options, 4/3 or 6/3 instead of 4/2 or 6/2? You only need 2 power wires plus the ground wire. If you use 4/3, you are wasting one of the conductors and making the wire stiffer and harder to work with than necessary.

Does anyone make 4-2 NM-B?
6-3 (on a 50 Amp breaker) would future proof for conversion to a 14-50 socket.

 

[dcgizmo]

Does anyone make 4-2 NM-B?
6-3 (on a 50 Amp breaker) would future proof for conversion to a 14-50 socket.

Honestly I haven't found it - so maybe that is the practical answer.
But for a 14-50 socket, I think it's wired differently and can't be used for a normal 14-50 application anyway? And the HPWC gen 2 manual says it has to be hard-wired; it can't be connected to a plug. The mobile connector can use a 14-5 obviously.

 

[mongo]

Honestly I haven't found it - so maybe that is the practical answer.
But for a 14-50 socket, I think it's wired differently and can't be used for a normal 14-50 application anyway? And the HPWC gen 2 manual says it has to be hard-wired; it can't be connected to a plug. The mobile connector can use a 14-5 obviously.

Sure, I was thinking of the case where the next owner of the dwelling has a non-Tesla EV and requires a 14-50 socket (with Neutral) for their charge cord.