Creating a low-cost EV charging station with Arduino
blog.arduino.cc
blog.arduino.cc
OpenEVSE is a properly developed ecosystem, and should be where you start if you want to actually do this. I cannot reiterate this enough; I find it mildly disturbing that the creator of this project fails to acknowledge it.
1. You probably should, Alec is great. 2. Here's a longer broader video about what goes on in an EVSE: https://www.youtube.com/watch?v=RMxB7zA-e4Y&t=1429s
There are too many comments to go sling context left and right, but generally an EVSE is a controller board (what this is implementing), a contactor (this board uses relays), and a GFCI (in North American parlance), which his RCBO provides along with being a current limiting circuit breaker.
Contactors are generally better for these use cases for reliability and current handling reasons but it wouldn't surprise me if SCRs and Relays are in use in low cost 16A EVSEs on Amazon.
I'm kinda puzzled about the system, though -- an EVSE is theoretically about as complex as a smart extension cable, but ... it stays warm even just sitting there not charging and gets to be 20 degrees F warmer than ambient when charging, which seems wrong for "just a relay"
My guess is warmth on the unit is a dumb inefficient DC transformer for the controller. There is an Energy Star spec for low standby energy use that most of these inexpensive ones don’t hit.
To anyone reading this who likes his content, I'm happy for you, I guess I just don't get it.
But he's not a subject matter expert on any of these things, and it shows. He's more like an independent reporter and/or journalist. Other channels on technical topics often hire experts to check them to make sure they're not making any obvious mistakes, and his videos would benefit from the same treatment.
* In the video about EVs not illuminating the brake light when regen braking, he confidently assumes that the problem is with all EVs, not just the particular brand of EV he drives.
* Not knowing that inverter microwaves exist, and confidently asserting that you can hear every microwave cycle on and off.
* (Over a series of videos) Letting his hatred for Tesla color his videos about EV charging.
For the last bullet: He went on and on about how every other brand uses CSS (with no acknowledgement that more than half of EVs in the US were, and still are to this day, made by Tesla). He ranted that CCS was obviously superior, argued that Tesla was trying to lock in their customers (despite EVgo and a couple others already offering Tesla plugs).
After the industry switched to the "inferior non standard", he started parroting the lines "well Tesla is just CCS anyway" and this is "Good For Bitcoin^H^H^H EVs".
It's too bad because there was actually a lot of good info in his EV charging videos. They're just so heavily slanted that I can't recommend them to anyone.
I don’t think the first point is fair tho. If this [1] is the video, at 1:27 he talks about two specific manufacturers. The point of the video was that regulations are outdated so different manufacturers can implement stuff differently, not that “everyone does it wrong”. Around the 13 minutes mark he speaks about GM doing the right thing.
Shortly after a test of washing machine tablets and powders came out[1], and none of the powders available there came out good in regards to naughty chemicals, while the tablets we had been using had none of those (Yes Platinum all-in-one).
They also came out better in regards to washing quality, so we've since changed back.
[1]: https://www.forbrukerradet.no/tester/vaskemiddel-til-oppvask...
For me, the results were unambiguous. With tablets I got a bunch of "water spots" that weren't solved by doing any of the typical advice like refilling dishwasher salt and rinse aid or cleaning the machine. The water here isn't particularly hard either. The machine also struggled with certain things like eggs. Pretty sure now that the spots were just from excess detergent, because it stopped happening once I tried the Technology Connections way. It also did a much better job with stains that the tablets struggled with.
Increasing washing temperature did help some but not enough. After a week my SO had had enough.
Perhaps it's related to the machine and water quality. We have very good and soft water (no salt needed), with very little chemicals. Perhaps another machine and/or different water quality can explain your different experience.
My machine is pretty old. Old enough that it has no electronics, just one of those mechanical dials that slowly turns back to "zero" during the cycle. There are no programs really, just the ability to set the dial to various subsets of the full cycle.
It wouldn't surprise me if more modern machines are designed more with tablets in mind than powder. Seeing as it's now the most popular product.
Yeah in comparison ours is a high-end Bosch model that's just a few years old. It has sensors which supposedly detect how dirty the water is and whatnot.
Would be interesting to do a comparison.
For us the main kicker was less of the troubling chemicals. We could have lived with the reduction in washing power had those metrics been inverted.
Better to comment if you have electrical expertise and can articulate a specific risk. Or perhaps maybe a sincere question toward the actual experts (guarantee there are plenty on this forum) about what the failure modes are (e.g. could this overload if it gets wet and the arduino malfunctions?).
However, there are plenty of gotchas in installing a EV charger, even a commercial off the shelf one. The regulations are generally pretty tight and are intended to protect from fairly unlikely scenarios. Getting earthing arrangement right etc.
Arduino DIY all day, every day... except for fire/life-safety critical systems like elevators, AEDs, automotive airbags or ABS brakes, car chargers, or large HVAC installations.
PS: I sympathize that the world hates almost everything fun or cool.
Commercial systems don't provide this (at least not without use of their proprietary and/or expensive hardware and software). There's nothing that says that you cannot design and build "fire/life-safety critical systems" that have open designs and open interfaces. Attention should be directed to these efforts.
All of the doomsaying in the comments here is ridiculously unproductive.
> All of the doomsaying in the comments here is ridiculously unproductive.
Not really. It might not be so much doom as practicality and legality. Reality is reality.
> All safety features required by SAE J1772, UL and NEC are standard.
(from the website)
Is it "UL Listed"? No. That's the end of it. There is no getting around that neither my city will permit this nor my electrician will install it no matter how much I protest that it's unfair. Sure, I could use it temporarily (not installed) or install it illegally without a permit but on sale of such a property, I would have to remove it to avoid it being discovered during an inspection.
If the important components and/or assembled kits passed Underwriters Laboratories' official tests and processes for listing, then my municipality wouldn't object and would take this seriously.
OpenEVSE v4 hardware (which I did advocate elsewhere) is UL listed and more closely matches with the goals of an open and accessible platform with a safe and tested design for those who wish to tinker with monitoring or software systems integrations.
Please stop being such a stick in the mud; its a boring way to live. "Reality is reality", bah.
how often are you buying and selling houses‽
It's designed to be safely used as an interface by third party hardware, and physically operates over a RJ12 connector (phone jack) with 5V. There are even DIY projects, powered by ESP boards, to interface with it - and electricity companies are ok with that, as it's designed for this use case.
This is no less safe than any other DIY installed EV charger.
Home insurance policies (in the US) often cover negligence including faulty electrical work. If it's in your contract, they will pay.
Insurance companies will then do their job, which is to find someone liable and send them the bill (subrogation). If you did the work, you're the responsible party.
TL;DR: Burning your house down is probably covered by insurance, but that doesn't mean you're not liable.
For example, DIYing electrical work in Australia is illegal. https://www.electricalsafety.qld.gov.au/electrical-safety-ho...
> Not only is it breaking the law, but you could also jeopardise your insurance!
> Changing powerpoints or light switches might seem simple, but unless you are trained and qualified, there are lots of risks you just won’t know about. Never attempt to do your own electrical work – you could kill or injure yourself or your family, or start a fire. Always use a licensed electrician.
Unless Australia has a much lower rate of home electrical accidents than rest of the world, I refuse to believe that this is a good thing.
Sorry, a bit offtopic but that partiular thing really grinds my gears, especially since it covers any cabling that goes inside your wall, so even just running an HDMI cable inside your wall requires a certified electrician to sign off on it.
Because yes, some homes are really vulnerable to electrical problems. And no, you can't trust any device (high-branded, non-name, DIY, whatever) on such homes. You can't even trust them to handle electricity being available, without any device consuming it.
A 10A outdoor charger probably also deserves it's own breaker.
You have a good source for this? I see claims like this every time a discussion of this sort and they rarely come with any citations.
Insurance will pay out, provided electrical fires are covered by your policy.
They will then do their best to assign liability and pursue subrogation. If you did the work yourself, you're liable. If you're insured for that work, they'll come after your insurer. If you're not, they'll come after you. If your work is permitted and to code, you'll probably be fine, if not, good luck!
So it looks like you're correct! Unlicensed work appears to be covered, but you may paid AND sued.
If you are not comfortable with high voltage electrical work, or are not familiar with basic components of the electrical code that governs your area, then definitely don't do this. But mass-labeling the work of others as unsafe or harmful is not appropriate either.
A smart electrical switch toggling dozens of amps over several hours... this might be getting outside the realm of what I trust cheap chinese amazon products with fake, misleading, or no UL certifications to handle.
So like two microwaves. It's also not running 20 amps through the Arduino, if the switch the software is flipping is rated for that much power then it's fine.
People suspect the traditional duty cycle used in rating these has been more for dryers, kilns, or welders, where they're active for an hour tops or at least not drawing full current. An EVSE can be at full power much longer, making the rating... less reflective of real-world conditions. I've heard there's a new rating system coming out designed especially for EVSE-sized duty cycles.
And that's not even getting into things like torque specs for the connections, etc.
None of this is blocking a dedicated, well-researched, and properly equipped DIYer, but at this level of power, there's enough "tricks of the trade" involved where the risk-to-reward ratio falls maybe outside the casual "buy things off AliExpress and who cares if the magic smoke comes out" range.
At this current, spending an extra couple hundred bucks might be worth it to have an actual company with a non-made-up-syllables name and a non-Shanghai-alley address that my insurance company can go after if my garage goes up in flames because their UL claims were misleading.
You’re probably thinking about DC fast charging, where the charger is external to the car and connects directly to the battery. Nobody should be trying to make _that_ themselves with an Arduono, sure.
Getting the amperage wrong or sending power when the car is not ready both sound like astoundingly dangerous failure modes for a DIY project, even if the most damage is bricking your car's battery controller.
The EVSE is talking to the onboard charger, the on board charger is what actually interacts with the batteries. The EVSE has a specific resistance inline with a comm pin to indicate presence (even when unplugged from the wall) and modulates a frequency on the comm pin to indicate max amperage. The car puts a specific resistance on a different comm pin to request that the contactors close to connect the charger to wall power. Any problems or bugs thereafter are with the onboard charger, not the EVSE.
Edit: my buddy is ex-googler. IQ is twice his weight. I wouldn’t trust him any electrical cable. He can work on ground breaking theories, but breaks everything what he touches with his hands.
Certain professionals should be licensed, because people need to know they are safe, but within your personal domain you should have a lot more freedom.
Because shit causes fires and burns my house down. That’s not to say it can’t be done safely, and if you’ve got the space to do it away from other people, such that the only person you hurt is yourself, then by all means I think you should be able to do your thing. But codes should come into play when it's being done in a garage attached to a home where the partner, kids, and dog live. if I blow myself up, that's totally on me. taking other people with me isn't okay.
As you said, the car side is dealt with by the car itself. There are some minor issues (read the YouTube comments, people on YouTube love to comment about safety, justified or not!) but overall quite reasonable from a safety perspective.
This is a slow charger and that car is not that different from a regular appliance from the charger point of view.
Not all of it; the whole point of having a controller on the wall box is that there's mains power in the cable only while it's plugged into a vehicle (and that this power is removed immediately if the cable is pulled out from the vehicle, which could help prevent arcing on the contacts if it happens while charging).
That is completely wrong.
The EVSE is responsible for implementing a GFCI cutoff as well as an over-current cutoff. Those are both safety features. A failure in the car's insulation resulting in a short to the frame is a shock hazard. A shock hazard also exists if the charge cord insulation is damaged. A bug in the car's charger firmware that results in too much current draw is a fire hazard.
Additionally, the EVSE must keep the contactor open unless the plug is attached to the car (sensed via the [edit: pilot, not proximity] pin). Again, a safety feature because the J1772 plug is large enough for a finger to accidentally contact the conductors.
Yet another is that the EVSE informs the car of the maximum safe current via a PWM signal. Too much current = fire.
The car can't help you with any of those safety aspects.
How is this low cost? J1772 chargers can easily be found for this cheap or cheaper. 16A chargers go for $120ish these days, and 32A capable ones are only... $140-150.
I don't know what you are referring too, maybe it's aliexpress level and they skimped on wire thickness. Most of the cost is the pure copper cable here, which should indeed cost 100+ for a 16ft cable, if you don't want it to melt when it's hot outside.
I don't know if they're skimping on the wire thickness, but frankly I doubt it. Tesla can offer their UMC for $250, and that's a branded product from a car manufacturer and is also capable of handling 32A charging all day.
It's a smart relay with a cable and industry-standard connector attached.
That's it.
It may have integrated ground fault protection
And the "smart" word covers over a lot of potential features, like stopping/starting charging based on electricity prices, doing load balancing between multiple chargers, reporting usage to the cloud, etc.
It's also an outlet where you really want to be sure that everything is professionally done, and up to spec, because drawing maximum currents for hours and hours is very demanding on all the parts. There can be significant heat generated in the wiring and contacts.
But that's not what the linked article is about.
I don't see a single thermal cutout on this design. I'd be scared.
The PCB is equally scary. How was the trace width figured for the high voltage side? There's also little or no isolation to the low-voltage side.
I also didn't catch if this is a level 1 or level 2 charger. Is that line 120V or 240V single phase? He's pulling 10A which may as well be level 1 household outlet charging here in the US.
The crimped ferrules on the wires is also much more common in Europe vs USA
https://www.gm-volt.com/threads/2016-volt-120v-evse-is-l1-l2...
I've confirmed this first-hand with several different models.
And it can be as low as 720 watts but I've heard of attempts to get it lower for certain situations where only 400 to 500 watts can be allocated.
Charging an EV in the garage, hmmm. Sure the probability of something going wrong is low, but it it does go wrong, it's a near 100% probability that the house burns down! So I'd rather have a possibly dodgy (but probably not; one can recognize sound enginering at a glance in the photo) charger and car outdoors, than one inside on a name-brand charger.
I have an EV, my house is old and its wiring is not very robust. But the garage plug can pull 6 Amps on 120V. That's 720 Watts. If I plug my car when I arrive home I'll get about 8 kWH overnight. On my Kia Niro that's enough for ridding 50 km / 31 miles. Add a little more on the weekend and it should be enough for daily commutes.
In newer houses with better wiring, you might get to 9 Amps and charge/ride 50% more.
Yeah, this alternative is really low cost, except the £80 RCBO [1] he uses for the DC current leakage detection. Something an EVSE does need to provide.
Oh, and also a full-featured EVSE will generally:
* Supply more than the 10A this provides, which you could get with a granny charger
* Monitor the current being drawn, cut power if the car draws more than negotiated, or more than the cables/relays are rated for.
* Have a reverse protection diode on the relays
* Use two double-pole relays controlled independently, so one always breaks the load (putting it at risk of damage) while the other never does (so when the first one fails, the backup is fresh and ready).
* Detect stuck relay contacts
* Sense temperatures on the board, at the main connectors, and monitor the release switch on the cable (if present)
* Manage relay coil power dissipation (a bigger problem with bigger relays, if your EVSE does more than 10A)
* More waterproofing - for example, conformal coating on the PCBs
* Don't put cable glands on the top face of the box, where water can pool up against the seal
* Don't involve a video showing someone poking wires into live electrical sockets, connected to bare circuit boards they hold in their hands.
Something tells me the person who watched this video has never read the specifications of a real EVSE, or watched a teardown by mikeselectricstuff or EEVblog...
[1] https://ecoharmony.co.uk/collections/6ma/products/bituo-tech...
On the other... hell no.