Inside the Hyundai/Kia V2L Connector

The new Hyundai Ioniq 5 and Kia EV6, both based on the shared E-GMP platform, support a cool “new” feature called vehicle-to-load. I use quotes because it is and it isn’t – plenty of cars have had AC outlets in them for plenty of years, and one 1.8kW inverter isn’t even that expensive on Amazon to just wire in (not that you could, on one of these cars, as I believe the DCDC converter that supplies 12V couldn’t handle 150A extra load).

On the other hand, the car already includes a bunch of power electronics to charge the DC battery from AC, and adapting those to run in reverse as an AC supply IS both relatively new and pretty cool.

Both cars sport an interior AC outlet, and both also support AC-out through their charging port, J1772 in the USA and Type 2 most of elsewhere in the world. As of now, the V2L connector is the one way to make use of the second option, and power stuff outside your car without running an extension cord through the window.

I mention that because, in lots of cases, it seems like a perfectly practical option. The only real points against are that you couldn’t well do that in the rain, and I can contrive at least one or two scenarios when I’d like to use AC powered devices even if it’s raining – a flea market stall, or camping with a tent, for instance. The main question is whether this device, which costs something like $500 in the USA, is worth it for that one use case. The second question is whether it holds any secrets as to whether we can get MORE power out of the car than just one outlet (your whole house, perhaps?)

So naturally I took it apart.

The first thing that jumps to mind looking at the inside is: I guess I can see why this device, produced at these relatively low volumes (a couple tens of thousand?) is super expensive. But on the other hand, it feels like it doesn’t NEED to be, and the design reads a lot like an automotive engineer got a sheet of requirements that felt too short to them, so they made some extra ones up.

For instance, look at that white shell.

That red o-ring on the right is waterproofing. It exists only to prevent water coming into the device through the power indicator LED window. But it’s a total farce for two reasons: 1. the opaque window is basically bonded fully to the shell. I’d be surprised if it isn’t totally water-tight on its own, and it could be made to be with ultrasonic welding, which is probably how it’s installed in the first place. And 2. assuming water DID leak in through the window, the o-ring would hold it in pooled contact with the LED circuit board, with no drainage path at all. That board has large, low-gauge AC wires exposed right on it. They designed in (expensive!) waterproofing that would make the failure mode MUCH worse than if the water simply dripped off that area into the rest of the device.

In stark contrast to the blue rectangle I’ve highlighted – a completely un-sealed, very-open leak path on the topmost surface, around the locking lever switch. This also leaks directly into exposed electronics (albeit at lower voltage on the neutral and proximity pilot lines).

It’s as if someone was like “this device is literally only useful in the rain, all exposed wires need to be hermetically sealed” but then someone else came along and was like “it’s impossible to hermetically seal the contacts in the plug” and then someone else said “it’s really hard to seal the lever we have to put in the USA version” so then they just gave up trying, but some expensive and questionable parts were already done.

Consider this expensive-looking definitely-automotive-grade hermetically sealed connector they’ve chosen to attach the AC lines to the region-specific outlet interface, which isn’t hermetically sealed and also can’t be.

Why does it even have a connector in the fist place? It’s true that there are a handful of regions this device must support with different outlets, but also the entire harness is different between the Type 2 and J1772 versions, so what’s one more thing with short wires you have to manually solder? It’s either manual labor on the connector, or on the other-connector, and it’s a totally overkill connector to boot.

Speaking of overkill, this is the LED indicator:

It’s a super straightforward design – there’s JUST a rectifier and capacitive dropper powering 3 LEDs directly off the mains voltage. If the output voltage is present, the LEDs light. Couldn’t be simpler. Although this PCB is desperately over-engineered for the task, with a test point for every net on it, and AC input coming in on like 14 gauge wires. I have no idea why these are so chonky – just to really boil any water that leaks into this cavity, I guess?

One last interesting bit of (I think) over-engineering is the thermal cutoff on the outlet itself.

The white ceramic piece is a bimetallic temperature switch that opens if it gets too hot, and in the assembly, it’s physically touching the brass contacts for the live and neutral (current carrying) contacts in the plug. The idea is, if the connection is bad and high-resistance and heats up to the point that the plastic might catch fire, this switch turns off the V2L output. Electrically, it’s the same as clicking the power switch off.

It feels a little silly to me because faulty outlets like that are caused typically by a combination of mechanical under-design and extreme old age. I think the outlet would be so annoying you’d replace the device long before it would catch fire. But on the other hand, maybe the real fault they’re designing against here is contact corrosion, on a device that’s basically expected to be used in the rain. Seems like a sensible safety interlock in that case.

In terms of overall function, I thought about drawing up a schematic, but for the most part it’s dead simple: the LED is powered directly by mains voltage present on the output when it’s enabled. The power switch and the thermal switch short control pilot and proximity together if they’re both “on” (i.e. cold). The latch handle switch connects proximity pilot to neutral via 75 ohms if the device is latched in place, or 500 ohms if it’s unlatched. And of course, control pilot gets the same treatment if the power switch is “on” since they’re shorted together in that case.

Region-specific output?

The most disappointing conclusion I have, in tandem with this British teardown and this test of a korean (J1772) connector putting out 120V on a USA car, is that nothing in this connector dictates what voltage the device outputs. It must be configured in the car itself. That seems to imply that a USA car won’t output 240V at higher current (like, say, 11kW) simply by plugging the right connector into it. There IS certainly possibility that more advanced communication supports higher-level output even with the hardware already on board. But this definitely precludes the possibility of simply making a passive adapter to let you, say, V2H during a power outage with the proper interlock on your panel and a generator plug (you CAN do this, but only for one phase, and only for 15A).

Here are some more pictures I didn’t include above: