DIY Home EV Charger

> Is it still a recommended practice on the US to rely on overcurrent protection to protect against current leakage? It's an horrible practice, as it requires several amperes of leakage before triggering.

Yes.

I don’t agree with it, personally — I think that there should be standard ground fault trip curves that are functions of circuit breaker rating such that they coordinate properly, and that they should be used at least on all circuits that extend outside of a building and really on most circuits, period. This would add costs, and it would detect and thus prevent a lot of dubious wiring practices and outright errors.

But you should still bond all your “ground wires” together so that even if these breakers fail to trip you have a degree of protection. (And breakers do fail.)

Under current NEC (the US code), when you supply a building with a feeder (i.e. from another building or similar structure), you must provide it with all the current carrying wires and separate ground, and you must ground it to the Earth.

In any case, if you live in a place where your utility supplies power above ground, and a utility pole or line is hit by lightning, its ground is most certainly solidly connected to your house. So you can’t really pretend that your house’s ground is connected only to the Earth near your house and that it won’t let lightning in.

Hmm, I wonder if you can take the ground wire running between two buildings and carefully insulate it and give it a wrap or three around a piece of steel to add some inductance and reduce lightning currents that might otherwise go through it. I’m not saying this is wise…

A lot of people would "shop" at brick-n-mortar stores with a mobile comparing prices online. They would look at the physical item right then and there, but purchase from online. At least, that was the early days of online when people were still actually getting off their couches.

Software projects have portability; you spend X amount of time, and others can use your labor for free or nearly free.

In comparison, hardware projects have costs even if the 'hard work' is done: PCBs to buy, components to assemble, parts to purchase. Even BEST case here, and discounting peoples time, it looks like the cost of all the components listed would be in the neighborhood of $200. And that assumes you can source the same 'demo' cable that the OP purchased for the same cheap price. Realistically I bet actual costs would be $300+

I have nothing against DIY hardware projects, the problem is that hardware projects must pass a much higher bar: you have to be cost competitive at a minimum with existing products on the market, or you need to be better in some way (and, that way can be trivial, as long as it's interesting). This appears to be neither.

The EVSE is a glorified RCD/GFCI cord with a contactor that turns on power, so the cord is dead until that happens. It also generates a signal that tells the car how much power it's allowed to draw; the EVSE is configured at install time to know what size breaker feeds it upstream, so it tells the car not to trip that breaker, but the EVSE has no means of actually enforcing such a limit. There's some basic negotiation between the EVSE and the car, to decide when the plug is inserted and it's OK to turn on the contactor.

The AC power passes straight through the EVSE, unaltered in voltage, current, or waveform. The EVSE may monitor the current to perform logging, and the EVSE is required to monitor the AC imbalance like any other RCD/GFCI does so it can open the contactor if there's a ground fault, but that's it. It's just a smart extension cord, basically.

The charger is in the car, and is the component that actually converts the AC from the grid to DC for the batteries. The charger is what has all the silicon (or these days, probably gallium nitride?), and the inductors and capacitors and stuff to actually transform the power from one form to another. It changes the voltage to match whatever the battery needs at the moment. It controls its own DC output to match the battery's needs, and it limits its own AC input according to the signal from the EVSE so it doesn't trip the upstream breaker.

The charger is also responsible for immediately turning itself off, reducing its input draw to zero, within a few milliseconds of someone pressing the latch of the EVSE connector handle. This indicates they're about to pull it out, and the contacts aren't designed to handle arcing, so the charger is required to drop its current _very_ quickly, allowing the contacts to be separated without creating an arc. That signal (called PP) is produced in the handle itself, there's just a tiny switch underneath the latch. The handle is technically part of the cord which is technically part of the EVSE, but the EVSE has no responsibilities related to the signal, that's all on the charger.

If the SAE hadn't mandated that the EVSE perform RCD functions, it could be replaced by a literal extension cord. Many small DIY EVs do precisely that; the charge cord is the exact same one you'd use to run any appliance. It works just fine. But even small DIY EVs have a real charger too, which converts the AC from the cord to DC for the batteries. The charger plugs into the cord, but the charger is not the cord.

How's that? Does plugging in a fridge require a permit? As long as the EVSE is plugged into an outlet like any other appliance, it's just an appliance. Your jurisdiction may require a permit for installing the outlet in the first place, but I've never heard of one where plugging an appliance into the outlet does.

Anyway, how many DIY EVSEs start fires? It's a glorified extension cord, not exactly rocket science, and as long as the builder applies a modicum of sense to choosing the components and wire gauges inside, and torques the connectors and stuff, there's really not much going on in there that should produce heat in the first place.

Also, this isn't a charger. It's an EVSE, there's no power electronics inside. There's a contactor, that's it, no silicon in the power path. It's really hard to screw it up.

Perhaps I'm just accustomed to the Hackaday comment section, but HN commenters are weirdly terrified of anything involving AC. It's just electricity. Not even very much of it, like, normal household levels of 240v AC. My clothes dryer uses more power, and it pulls it over wiring installed in the 1970's. Heck, when I moved, I even replaced the cord on my own dryer with one that matches the outlet in this house, an action explicitly allowed and encouraged by the access plate on the back of the appliance. The bare cords are sold at Home Depot and they come with a little slip of paper explaining how any homeowner can install them.

It's fine. It's just wire. Pick one that's thick enough, torque the screws, it's fine.