5-20 dropping 13 volts under 16 amp load. Not good.

[timk225]

I have the opportunity now to charge from a 5-20 outlet at work. I was happy for this, they even checked and replaced the GFCI outlet for me. It's in a 6 year old building that is designed to use a lot of power for computer stuff, and as near as I can tell from looking at the breaker panels, and the outlets are all marked for what breaker they go to, there are only 2 outlets on what should be a 12 gauge wire. The wire run is about 130-150 feet, maybe a bit more. The other outlet has nothing running on it for power draw.

When I plug in my 3, the voltage starts at 120, but by the time the charging ramps up to its full 16 amps, voltage sags to 107. It's right on the edge of tripping the charge off. I initially used a new 12 gauge extension cord and was only able to get 105-106 volts with constant extension cord warnings.

Today the car charged at 107 for a while, then it had the charge going on and off, up to 120, back to 107, pausing in between, etc. So I dropped the charging amperage from 16 to 15, and that stabilized it at 108 volts.

There's only 2 GFCI outlets on this circuit, and I've traced the wires. Initially I thought maybe the wiring to my outlet was going into and out of the other GFCI, but I pulled it out of its wall recess, and no, it only has a single hot, ground, and neutral going to each of the 2 GFCI receptacles. So there's a Y connection somewhere in the system, and it's all in metal conduit. So it'd be a real pain to pull it apart and look for a bad connection.

Or maybe I am over thinking it, and it's just a long run from the breaker panel even for 12 gauge wire causing the voltage drop.

 

[Dmagyar]

Voltage drop is a product of any amperage drawn, and length of a particular wire gauge (size) as distance and amperage goes up, so does voltage drop. Your extension cord adds to the overall voltage drop, the condition and age of that cord has some bearing also (old, and with some strands possible broken along the length). A new #3/12 cord shouldn’t appreciably affect drop unless it’s over 25’. The only way to combat voltage drop is with larger conductors or a shorter run or less amperage drawn. It also may indicate that the voltage available at the panel or source to be low, because of combined loads to said panel and the diversity originally planned for. If the building owner or utility is contacted they may be able to check for the voltage drop at the panel and remedy that by changing the taps at the supplying transformer. Good luck

 

[Sandor]

Kinda excessive drop. Really fast back of the envelope calculation
12V drop at 15A that's .8 Ohms worth of wire
12 AWG is about 1.7mOhm/ft under load (I assumed 50deg C; given the180W loss that should be on the low end)
That makes it 470ft of cabling from the breaker to the car. I doubt that is that much.

Check the panel or the wiring something is fishy somewhere. Given that you have 180W going somewhere a IR detector should pinpoint it as long as it's not in the conduit underground. But in the conduit I'd expect stuff to have caught on fire by now as those 180W wouldn't have anywhere to go ... Check the panel and check the outlet first. It may be very well that the panel puts only 107V under load - under-dimensioned for the load

Hope it helps,

 

[Sophias_dad]

Sandor, the 1.7mOhm/ft is for a single conductor, so your 470 foot distance is off by a factor of two, I believe. That's 235 feet of two-conductor cable involved in the circuit.

 

[Sophias_dad]

what should be a 12 gauge wire

Per Voltage Drop Calculator | Southwire, a 140 foot run moving 16 amps should have 8 gauge wire to attain a <3% voltage drop(3.094 volts, specifically). Doing the same with #12 gets me 7.77 volts of drop.

I'd agree there's probably something else amiss in the circuit, but you really gotta check the panel's input voltage and circuit breakers output voltage under load. It would be entirely reasonable for the power company to deliver you several percent less than 120V continuously.

https://www.pge.com/includes/docs/p...ergystatus/powerquality/voltage_tolerance.pdf might help here.

 

[timk225]

I did some more testing on the circuit today. When initially plugged in, the car was showing 122 volts on the screen, before it began to pull any power.
At 5 amps draw, it showed 118 volts.
At 10 amps draw, it showed 114 volts.
At 15 amps, it was at 109/110 volts.

I put it back up to 16 amps briefly, and it showed 108 volts. I can't imagine why, but I have to wonder if there's some 14 gauge in the circuit somewhere. I traced the wiring as far as I could throughout the building yesterday, and it was in metal conduit all the way. It's definitely 12 gauge at the GFCI receptacle.

Power input to the building is 480 volts through a 1500 KVA transformer (currently reading 489 volts input), then everything in the building is either running on 208 or 120 volts, depending. I'd pull the cover off the breaker panel and check voltages, but they won't allow that.

 

[Frank99]

It was good that you noticed the voltage drop. Most would have blown it off.
Currently, you think that the circuit is "130-150 feet long". Assuming that you paced it off, that's the minimum. You haven't fully determined the route of the circuit ("So there's a Y connection somewhere in the system"), perhaps it's bit longer. A compliant circuit won't have a Y stuffed into the conduit somewhere; it would have to be in a junction box, and the most likely place is in an outlet box. Perhaps you missed one when you were looking?

Anyway, you're seeing a 13V drop.. Using the relationship Resistance = Voltage/Current, Let's do some calculations from your numbers:
13V drop at 16A, R=V/I = 13/16 = 0.81 ohms.
12.5V (average of 109/110V readings) drop at 15A, R=12.5/15 = 0.83 ohms
8V drop at 10A, R= 8/10 = 0.80 ohms
4V drop at 5A, R=4/5 = 0.80 ohms
So we have very consistent circuit resistance measurements. Now, how long would the circuit have to be to show a 0.8 ohm resistance? Using the Southwire calculator data of 0.0019029 ohms/foot for AWG 12 wire, we have:
0.80 ohms / (0.0019029 ohms/foot) = 420 feet of wire = 210 foot circuit.

Perhaps a more detailed tracing of the circuit will find an overlooked junction box whose position adds 60 to 80 feet to your estimated circuit length?
I don't think you have a problem here; you just have the physical limitations of wire and Length. To make the car happy, you're gonna have to charge at reduced current.

 

[Rocky_H]

Per Voltage Drop Calculator | Southwire, a 140 foot run moving 16 amps should have 8 gauge wire to attain a <3% voltage drop(3.094 volts, specifically). Doing the same with #12 gets me 7.77 volts of drop.

I did some more testing on the circuit today. When initially plugged in, the car was showing 122 volts on the screen, before it began to pull any power.
At 5 amps draw, it showed 118 volts.
At 10 amps draw, it showed 114 volts.
At 15 amps, it was at 109/110 volts.

I don't think there is much that is noticeably off about this. It's just a really long circuit, and someone didn't bother to upsize the wire for that. As pointed out, the 12 gauge wire is going to have about 8 volts of drop in a best case anyway. I would not at all be surprised for any number of reasons to see that about 10 or a little over. I would be curious how the wires are attached into the outlet. If they are using those "stab-ins", where the wire just pokes into holes in the back, that is a notoriously bad tiny contact area connection. That will add significant resistance. It should instead be using the side screws, where you can curl the wire around it and then screw it down to make a tighter, more full connection.

But as you noticed, if you just turn the amps down from 16 to something like 12 or 13, it cuts the voltage drop quite a bit. That's what I would recommend. Someone should have used a little thicker wire, but didn't, so that's what you've got.

 

[EVDRVN]

I would drop the amp rating for safety. You can have an unknown junction or high-resistance on the breaker or elsewhere causing the drop.

 

[timk225]

I would be curious how the wires are attached into the outlet. If they are using those "stab-ins", where the wire just pokes into holes in the back, that is a notoriously bad tiny contact area connection. That will add significant resistance. It should instead be using the side screws, where you can curl the wire around it and then screw it down to make a tighter, more full connection.

I checked that myself on the GFCI 5-20 outlet. The wires do go straight into the back side, but they aren't the stab-in type. The wires still go under the screw and little square plate and are held down by clamping force. I feel that's not as good as truly wrapping the wire around the screw, but certainly not as bad as just a few tiny little metal teeth points having all the power flow through them.

The car has been on charge now for several hours at a steady 15 amps and 110 volts. It was reading 111 volts for a while but is now back to 110.

Here's the exact GFCI outlet -- GFCI Receptacle,20A,125VAC,5-20R,White HUBBELL WIRING DEVICE-KELLEMS GFRST20W A5

 

[Frank99]

A clamp style wire termination on the outlet is an excellent termination; no need to question that as long as the screw is tight.

 

[timk225]

Yes, I shut down the power to that outlet, and checked the clamp screws, they were tight. I think the biggest issue is just the length of the wire.

 

[Dmagyar]

If only more electricians had access or had in their possession an Ideal circuit analyzer 61-165 or the current model, it easily lets you know the condition of any branch circuit, and it’s potential voltage drop at both 16 and 20 amps of load...plus circuit wire resistance all without unscrewing a cover or plate...

 

[BogStandard]

I know this is a bit of an old thread, but I saw on the "similar threads" suggestion at the bottom and thought it was interesting.
If those outlets won't ever be needed for anything else, or isn't required by code to be there, that circuit could be swapped out to a single 6-20 or 6-15 receptacle.

The breaker would need to be changed to a double pole GFCI though, not sure if there's enough room in the panel or is allowed by your building owner.
Then you could just buy the appropriate gen 2 adapter and have 240v charging (or 208 most likely in your case). Even at 12A on 6-15 circuit, you'd be getting ~9mph of range added.
Although, without looking it up I'm not 100% certain you can put a 6-15 or 6-20 receptacle on 208v, I have a feeling maybe not.

But hey, even at 110v, getting 20+ miles of range throughout the day for free is something. It at least lessens the battery draw when you want to pre heat or cool.

 

[Rocky_H]

Although, without looking it up I'm not 100% certain you can put a 6-15 or 6-20 receptacle on 208v, I have a feeling maybe not.

Yes, definitely. The 240V or 208V are treated pretty much interchangeably for all that category of outlet types.

 

[M3BlueGeorgia]

I did some more testing on the circuit today. When initially plugged in, the car was showing 122 volts on the screen, before it began to pull any power.
At 5 amps draw, it showed 118 volts.
At 10 amps draw, it showed 114 volts.
At 15 amps, it was at 109/110 volts.

I put it back up to 16 amps briefly, and it showed 108 volts. I can't imagine why, but I have to wonder if there's some 14 gauge in the circuit somewhere. I traced the wiring as far as I could throughout the building yesterday, and it was in metal conduit all the way. It's definitely 12 gauge at the GFCI receptacle.

Power input to the building is 480 volts through a 1500 KVA transformer (currently reading 489 volts input), then everything in the building is either running on 208 or 120 volts, depending. I'd pull the cover off the breaker panel and check voltages, but they won't allow that.

IMHO you should consider restricting your power draw to 12 Amps, or maybe 14, but no higher.

 

[Dmagyar]

IMHO you should consider restricting your power draw to 12 Amps, or maybe 14, but no higher.

Did a quick calc on my voltage drop calculator, length shows up at 200 feet with that drop and #12 wire, but remember it’s the total length of both the hot conductor going out and the neutral coming back to the source for the calculation… Extra resistance can occur in splices also, just something else to note.