The Charger Games:   scary stuff!

  There was much to be explored in terms of charging the car.  It is important to understand what options are available and what their limits are, because not every recharging situation may be in a comfortable known environment.  What do you do, for example, if you're visiting friends for a weekend who are just beyond round-trip range and/or you want to do some driving around their area while there?  Or that urgent mission to the middle of bumfuk District 13, and you want to be confident you can get the hell out of there before the Peacekeepers show up?  Being able to charge in an optimal way at a destination is important, which is why it's called "destination charging" and is often associated with an existing installed EVSE or high-power outlet.  But we don't always have that, and the wheezy little "granny charger" that comes with the car may not be enough to support local activities and/or getting back home in a reasonable timeframe.

My Tour de Sol buddy up north had seen rumors that the small EVSE that came with his Bolt could run on 240V and double the charge rate of the car.  This sounded entirely reasonable, as it's really just a relay box with some support and interface stuff added.  Modern low-voltage switching power supplies, such as for the electronics to handle the J1772 protocol and energize the relays, often have wide ranges of input voltage so that they can run in US or European markets unmodified.  Further supporting evidence can be found in this long-running thread at insideevs.

I briefly looked at cracking the lid off the glued-together box of mine to look at the innards, but decided that doing so would probably be fairly destructive and that there were other ways to externally and safely test this theory without blowing the thing up.


Scary charger test lashup So here's my collection of scary-ass charging widgets, in the test setup to determine the capabilities of the granny-charger.  The "combiner" is fed from the wall outlet and a cord run from another one on the opposite leg.  From that outlet emerges the key piece to doing this at all -- a totally redneck adapter thrown together from NEMA 14-50 to a 5-20 Edison outlet, presenting 240V across its hot/neutral and ground at the ground.  That's the short bit with the multicolor wires, yellow ends, and exposed live terminals.  I left all the covers off that piece for a reason -- it's *supposed* to look weird and dangerous, because it's only for this single purpose.  What's then plugged into that is a "series outlet" setup on the grey cable, with the two outlets of an Edison duplex un-ganged and then chained together neutral-to-hot across the middle.  That provides a place to drop in a "dummy load", consisting here of two 120V light bulbs in series, to protect whatever's in the other outlet.
And into that other outlet went the EVSE plug.

The light bulbs, all of 25W in this setup, came on rather dimly, and the EVSE powered up in what seemed a perfectly happy fashion.  So I was already fairly sure it would be fine on a 240V feed.

I'll point out that the series-bulb setup is a time-honored way to test devices that may be internally shorted, such that a repair tech can apply power to it safely and probe around to find the problem.  The light bulb(s) can be sized appropriately to the device under test and the desired current limit.


Charging tests with series load Next step was to drop in bigger light bulbs, a pair of 200-watters, and repeat the power-up test.  Still fine; the bulbs didn't glow at all this time, meaning just about all of the 240V was across the EVSE.  Now it was time to take the whole mess out to the car and give it an honest go from my real feed outlet.

What I observed next was rather fascinating.

After I plugged into the car, the EVSE's car icon went blue to indicate "charging", the relay clicked on, and the bulbs began glowing -- a little at first, but becoming brighter over the course of a few seconds.  The car's onboard charger was clearly doing a slow current ramp-up, doing its best to avoid any inrush and be gentle on its supply feed.

Charger error, stopped Of course the light bulbs prevented drawing any meaningful current, and the 240V to the EVSE rapidly collapsed.  Relays klunked under the hood again and the car started blinking red in the ring around the J-plug, indicating "charging error".

I repeated this experiment a couple of times, both to get pictures and to get a sense of the ramp-up curve.  It went in two or three gentle steps over about as many seconds before failing, but it wasn't a smooth linear ramp.


Granny charger works fine on 240V Anyway, at this point I knew the EVSE was fine and it and the car were going through the whole protocol dance, so I removed the series bulbs, plugged straight into my scary Edison adapter, and went to town on a real charging test.  Set to 12 amps, the "granny charger" was happily feeding the car twice the power it would have off a regular household outlet.  Case closed on this EVSE, and I was more confident that my buddy's Bolt bump-in-the-cord would be fine with a similar setup.

Real maximum OBC charge rate Having only recently gotten the car home at this point and mostly-finished running the real 14-50 driveway outlet, I hauled out the GoPlug for a quick test as well.  It delivered about what I expected, even better than the spec "7.2 kW" stated in the literature.  I was probably giving it an honest 240V, not 230 or the like that other areas may have -- my power company has done a lot of solid upgrades in the neighborhood over the last decade or so.  The power that the car draws is very dependent on leg-to-leg input voltage, which is also the usual reason charging stations at commercial properties with three-phase 208/120V power can only deliver a little over 6 kW at 30 amps instead.

I soon stopped this, though, as I was still on the "dealer charge" and intended to run that all the way down to nothing before recharging for real.


Next experiment was the potential for "rescue charging" on the road, using portable power packs I'd recently acquired for outdoor events.  This *would* be a 120VAC scenario and thus slow, but each kilowatt-hour you could bring to a stranded EV could potentially get it four or five more miles down the road.  Thus, perhaps these boxes could be like a very tiny gas can in the trunk.
Yeti pack with no ground: fail Yeti pack with faked ground: works!
There was a problem, though: these packs have a big empty hole where the ground pin of an inserted cable fits, and no ground connection.  An EVSE needs a solid ground for fault detection and the pilot reference, so this wouldn't work as issued.  The Yeti pack output is fully floating, however, and either side of it can be connected to a "fake ground" to simulate a normal grounded outlet.

The small wire hanging out of the floating connection in the first picture is a copy of the "neutral" lead, or the lefthand prong of the box.  With it floating free, the EVSE indicates a ground error.  With the small wire connected to the shell of the outlet, thus creating a ground equivalent to the neutral lead, the EVSE is now happy.  The fake ground wire can be tiny, because it's only a reference and will never carry much current.


Yeti power-pack can charge the car I then hauled that whole rig out to the car for a test, and the Yeti pack could charge the car at the full 12A setting on the EVSE -- maybe not quite 12 amps for real through this lashup, but the Yeti boxes are quite solid and can deliver an honest 1500 watts steady-state as advertised.  So in theory I'd get close to its rated kilowatt-hour out of it top to bottom, less minor losses in the charging chain.
As I said, it's nice to know all of one's charging options, and by going through all of these tests I had functionally explored several of them and put together the right adapter widgets to make it possible.  To summarize:

  • "combiner" from household outlets to GoPlug -- 3.6+ kW

  • granny-charger on 240V -- 2.8 kW

  • granny-charger on rescue-pack or wall outlet -- 1.4 kW

  • GoPlug on its real feed -- 7.2 - 7.4 kW, limited by the car and/or supply voltage

and that doesn't cover any of the high-power public charging options.  Rapid charging is extensively explored in a subsequent section.

If you stop and think about this, that's still quite a bit of power we're casually talking about here.  What household appliance normally draws over 7 *kilowatts*?  Dryers and water heaters are usually 5 kW or less, and intermittent.  Maybe a big hot-tub or wall-mount heaters in an all-electric home?  I already knew that home charging would probably double my electric bill or more, but it's still much cheaper per mile than gasoline, even in a competently-driven Prius.


  In search of more cross-compatibility

  There are various forum tales of people needing to charge a J1772 vehicle, such as overnight at a motel on a roadtrip, but arriving to discover that there are only Tesla-style "destination" charge points available.  There's also plenty of information on how to get around this with adapters, which have evolved from the earliest homebuilt "Tesla tap" lashups to convenient commercial products.  As long as I was building a "road kit", even with a relatively remote possibility of myself ever needing to charge at a Tesla point, a few more months into ownership I decided to pick up one of the adapters that were now commonly on the market.  The prices on such things were clearly coming down; I went for the Lectron product at far less cost than what the others were demanding.

The Lectron Tesla destination-charger adapter This is it; looks simple enough, just changing one plug into another.  Seems fairly solidly made, and conveniently short to throw in with the rest of the widgets in the back boot of the car.

Inside the Tesla head of the Lectron At least one of the other products claimed that active electronics were needed to "fake out" the Tesla-side EVSE in some way, because it defaults to some Tesla-proprietary protocol before falling back to J1772.  Other sources say that's false, and that the Tesla box would simply time out by itself and start charging whatever's on the other end.  There was some muck about internal DIP switches to set on the destination chargers to make faking it out easier or harder -- well, none of that would be relevant here, because opening up the Tesla head [which isn't easy] shows it to be a completely passive pass-through for AC, ground, and the pilot line.  The J-plug at the other end only has the standard proximity resistor ladder.  Level-2 Tesla charging theoretically *is* J1772 at its heart, it's just a question of the different connector.

Looks like the plugs will match So I really had no idea if I would run into difficulties with this, and the only way to learn would be to go find a Tesla EVSE and try it.  I hopped on Plugshare to search around, and the nearest public one appeared to be in Methuen.  On a bleak day, bleaker still as the gravity of the coronavirus problem was starting to become clear to everyone, I tooled out there to give it a whirl.  It wasn't occupied, so I pulled right up for a look.  It's actually a dual-head post with two destination chargers, and another one with two J1772 heads near it serving a total of four parking slots.  And the post had power, shown by the green standby light.  On examining the Tesla plug and my adapter, it seemed likely that they'd fit together without issue.

To Tesla's credit, they did design a fairly sleek and capable connector type, with quite a large contact area on the high-current parts.


Hey, it works to charge the car It took a couple of tries, mostly because I was slow about letting the latch down so the Kona could grab the plug the first time, but the car soon made the right noises, turned the LED ring green, and began charging.  The descending "green worm" on the Tesla box confirmed that it was fine with things.

The slow let-down was on the vague theory that letting the Tesla see the presence of the car for a few moments before actually requesting charge would help get around the supposed EVSE timeouts.  The opposite was true, actually -- when I just plugged in and let go, *that's* when it began charging correctly.


6.9 kilowatts, more than typical commercial 3-phase The EVSE was delivering even more than I would expect from a commercial feed, which is usually 6.2 kW from a 208V leg-to-leg supply in such locations.  Perhaps the voltage was a little high in the area, who knows.  Test concluded; I didn't need to hang around for any length of time.
A mildly hilarious note, as I was doing this ... a worker came out of the Chick-Fil-A nearby and came toward me, saying "that's only for Teslas!" and waved at the pair of Clipper Creeks on the next post as where I should be.  She was trying to help me out, not chase me off, evidently thinking that I might just be confused about charger types.  When I told her I was testing adapters she was like "oh, okay" and went back in, as she clearly had more important things to attend to.  Restaurants had just gone into "take-out only" mode over pandemic concerns, and the place was absolutely slammed with a very long line for the drive-thru.  I'm surprised she even took those moments to pay attention to me at all.

  [ Up to EV index ]

_H*   191003, 200325