A circuit breaker designed for electric cars to lower costs

A shiny new Tesla, with 15 miles on it

While I was watching a rocket lifting some of my friend's satellites into space from my driveway yesterday, my new electric Tesla car was delivered to that driveway.

Now that's a sentence that says you're truly in the 21st century. The SpaceX Falcon9 was launching 64 different micro satellites, some of which were from my friends at Planet Labs. What used to be the province of countries is now done by the startup down the street. And it was just an extreme coincidence that the Elon Musk Tesla arrived at the exact same minute as the Elon Musk SpaceX launch. You do have to hand it to Elon, even with the differences of opinions we have on LIDAR and other issues.

If the opening sentence sounds like it says "look at how fancy I am" (and it does) consider:

  • The friends are not billionaires. Cubesat lauches are in the tens of thousands of dollars, not the millions.
  • The Tesla Model 3 medium range model without autopilot is $46,000. After $10K tax credits it's $36K. After gasoline savings (if you get them) it's $32,000 -- and it's a really great car for that fairly ordinary price.
  • Anybody in most of California can see launches from Vandenburg from their driveway.

So you'll see more posts from me about electric cars in the near future, I'm sure, so let's start about the question of home charging. In the robocar world, cars will be able to drive off to find charging when they are low, so you won't need to install home charging, but today you want it.

This means having some sort of "charger" with the big car charging plug. We call it a charger, but in fact the charger is in the car, the home unit is just an interface between the car and the grid which tells the car about the power available. It is surprisingly expensive considering that's all it does. Wall units tend to run $400 to $600 (and even more.) Tesla's mobile unit is a surprisingly cheap $300 compared to those. Almost always, you will need to run a new dedicated high-current circuit from your parking place to your power panel, and for most people that means a professional electrician install, possibly even with a city permit, and you can hit $1,500 in costs without much effort.

Contrail of rocket from my street, 1 minute after Tesla arrived

But worse, older homes that only have 60a or 100a of service may need to do a full panel and service upgrade. That's on top of the EVSE (the real name of the "charger") and can readily run $4,000 or more here in California. That pretty much uses up several years of the money you planned to save on gasoline with your electric car. (And I do want to stop burning gasoline or buying anything from Mohammad bin Salman. In California, my energy company offers a 100% renewable option that is cheaper than the non-renewable price from PG&E.)

Some cheaper alternatives

Some lucky people may already have a spare high current circuit or plug in their garage or near their parking, most commonly the one for their clothes dryer, or sometimes an unused one for an electric range. If you don't use those (for example you have a gas dryer or range) you might be able to just use them or put a very special extension on them. And some people sell special switches you can put in these sockets which split out into two plugs, but turn off the car socket when the dryer is running. This assures your house never uses more current than it originally would. These plugs are 30a, meaning they can charge a car at 22 amps, which is slower than the higher end cars can handle.

Smart sharing of power

This led me to an idea for something that should exist. There are partial forms, but offerings could be a lot cheaper. The car is smart, can control its current, and has wifi. The EVSE may not be that smart, but has the ability to interrupt the current and tell the car how much current it can output. And in the main power panel, there are many things, including the circuit breaker for the EVSE which also has the ability to disconnect the power, and there's also a smart meter which knows how much power the house is using, and the ability to put loops around the conductors there to know how much current the house, or any given circuit is using.

Instead of putting in extra, expensive capacity, the smart devices should be able to learn how much spare current is available at any given time, and only draw that much current.

Here are a bunch of ways that could be done. Some are easier to install, some are cheaper, some give better results.

Smart meter and smart car

I have a smart meter. You can connect to it over Zigbee Smart Energy and get a stream of information on real time power usage. The Tesla doesn't have Zigbee but it could talk to an inexpensive gateway on the home wifi and get this information. Then it could track house energy usage, and decide when to charge. If other appliances are drawing lots of current, the car could stop or scale back. It would leave a large enough margin so that any one appliance (oven, dryer, air conditioner, hot tub) could get turned on and not overload, and then it would scale back so that any new device would not overload. If the stream of data stopped for some reason, it would stop charging. If it stopped for long, it could send an alert to fix things.

One problem -- a car visiting your house would not know to follow this protocol and could overload the panel. UL probably would not approve that. So in the end, while the car is nice and smart, some change to the EVSE is needed. This could be a fully smart EVSE, described below, or a minor modification that it does not offer power to any vehicle that doesn't identify that it knows and follows the protocol. Tesla has a protocol to negotiate with their EVSEs, but other cars don't have this.

Smart EVSE

The more correct way to do this is to put these smarts in the EVSE. Let it learn what load is present in the house, and adjust the current it offers. It could learn that from the smart meter -- it could speak Zigbee Smart Energy protocol.

Current measurement at the panel

It's easy to measure current in AC wires. You just put a small loop around the wire and it induces a voltage. Electricians could install these loops. They could either feed to the EVSE over some extra wire put in the conduit at the same time the big current wires are run, or it could tap off a little power to run a digital device which would read the loops and send the value by RF or signal on top of the powerline. This signal would get to the EVSE and let it know the current with no smart meter.

Current measurement at appliances

For those unable or unwilling to measure current at their panels, a small device could be made which would be built into a unit which would be inserted into dryer or range plugs. For example, you could get a unit that plugs into your dryer plug, and then the dryer plugs into it. The unit would read the current used by the dryer and transmit it, via RF or signal over powerline, to the EVSE (or the car.) The EVSE would now know when the dryer was on, and signal reduced or no power available while this is so.

A special circuit breaker

It would be most elegant perhaps to put all this into the large circuit breaker which controls the car charging circuit. That's because that breaker is, of course, in the power panel box where it's easy for it to run loops to measure the current on other devices there. It's also because that breaker has the ability to cut the power on the main current lines to the car, eliminating the need for the EVSE to do that. (A proper EVSE offers no current until the amount has been negotiated.)

This would mean the EVSE would be almost nothing. It need only generate the pilot wave to tell the car how much current to use. It could perhaps not even have to generate that. A 5 6 wire convention could be developed for the run from the power panel (and smart breaker) to the EVSE, which would really be mostly just a pass through to the charging connector.

The circuit breaker could also enhance safety by cutting power if the car does not reduce the current when it is told to.

Smart load management

Almost all the big loads in a house except the electric car are actually intermittent. Air conditioner and refrigerator compressors, oven heaters, dryer heaters etc. all turn on and off. Likewise the car can take current in the low cycles and refuse it in the high cycles. Pool pumps often run all night, but new modern pool pumps usually are multi-speed, and run at low power (just 2-3 amps) during their night circulation mode.

Getting fancy

As noted, the system could also put current measurement rings on the big loads. This would let it learn their maximum load, and know which particular appliances are turning on and drawing power, to help it adjust the current.

Some day we'll have smarter panels that do this for all circuits. And it makes sense for all appliances to talk a protocol like this to manage power. A house where all the major appliances listed had a system to "take their turns" could get by with a vastly smaller connection to the grid. And by evening things out, they would be doing the grid a favour. The hot tub doesn't care when it heats (unless you are in it.) The pool doesn't care when it pumps. The air conditioner and fridge can wait a few minutes to cool if needed.

This makes the charging point electronics very simple, and much cheaper. In fact, I believe this design would eventually make the whole install cheaper once it is made in quantity. The main downside is you must make breakers for all the major panel designs if you want to use them to interrupt the power. Code requires there be no voltage in the main pins until the handshake has been done with the car. The circuit could also go in a small box clamped to the side of a main breaker box.

But even if this circuit cost a lot extra, as it might to start, it would be far less than the cost of upgrading your whole house electricity, which is something many will face as they buy electric cars.

Do you need Level 2 charging?

Home charger installation can get quite expensive -- as I will detail below, it can involve $5,000 or more for many people. But as the batteries get large, it is no longer necessary to keep a car fully charged, unless you know you are going on a road trip the next day.

This means that it is practical, but not as nice, to charge with an ordinary household plug with 15a or 20a max. These are plentiful, though you want one that is on its own circuit. They charge very slowly -- anywhere from 3 to 5 miles of range are gained with an hour of charging.

But the average car in the USA is driven only 32 miles per day! My 220 mile range Tesla could do a week of average charging on one charge, and would do just fine ordinarily gaining only 50 miles each night from the wall plug.

The problem of course is that cars don't drive exactly 32 miles every day. Some times I go to San Francisco (100 miles RT) two days in a row, and this would be barely doable and put me in the "range anxiety" mode. So this is where fast DC charging can save the day. It can mean a 30-50 minute pause, eating a meal at a supercharger, but it probably only needs to be done a few times a year. It might be a much simpler alternative than adding expensive high current charging in the home.

One problem is that if you use only 12 amp charging, you are very tempted to charge not just at night. Here, the electricity costs 3 times as much in the day as in the night if you have the EV plan. That gives up a lot of the savings you might have hoped for.


Developped in EV friendly Quebec, and commercially available since this year...


See DCC-10 product too.

It is a bit pricey -- but yes, that is still way less than a panel upgrade. It just cuts the power (and keeps it cut for 15 minutes after the other load came in.) But in theory in most houses it would never actually cut, not at night. The highest load in a house otherwise would come if you had the AC running, food in the oven and were drying a load of laundry.

I use variablegrid, can add two 40A circuits for charging to a 100A panel.

I use variablegrid, can add two 40A circuits for charging to a 100A panel.

We got the DCC-10 for this purpose, and it works like a charm! We're able to deliver 48A to the car without having to upgrade our 100A service, and in a way that passes inspection. It saved us several thousand dollars in a service upgrade. In the six months we've had it, it has only cut power to the car twice (to handle the load of a couple of late-night clothes-drying sessions), and only for a half hour or so. I highly recommend this product!

Welcome to the Model 3 club, Brad! (Geesh, you write my favourite C64 assember, go to the same university as me, write my favourite obfuscated C program, follow me to Burning Man, buy the same digital cameras as me (participating in the same discussion forums), and now you buy a Model 3? Crazy!)

I am sure there are many others in the same pattern. That DCC-10 is a start but since most of these EVSEs contain a microcomputer, they really could do a bunch more smart stuff. Other appliances should join the game too. As I've gotten to know the Tesla I have been surprised that it does not get smarter about this. (Though I am not sure what the UL and code inspectors would think about the Tesla being the smart device and not drawing power when other things are.)

But it doesn't even have a setting to stop charging when peak prices begin (there are apps to do that) or otherwise understand anything about ToU but when to start charging.

Hi I've had a Model S for six years and just charge it from a standard 220V / 30A plug that was already in my garage for an electric dryer. It works great. Of course it doesn't charge as fast as a 50A circuit or a dedicated Tesla charger, but it's fine for charging the car overnight and cost $0.

The most important 'fact' and aid to EV adoption is the introduction of true fully autonomous Self Drive Vehicles (faSDVs"). Absent faSDV tech, EV will have significant adoption issues:
[1] half of all families earn below $60,000 per year and have less than $1,000 in the bank. They cannot spend $5,000 on in home charging and even $1,000 can be difficult.
[2] in urban areas, residents might have NO personal parking spots. Just street parking. No place to plug into a home charger.
[3] in urban areas, no government has the surplus funds (and voter support) for high density street charging. To bring street charging to a thousand 1 mile street segments in LA might take longer than a decade and cost billions.
[4] in suburban areas, high density private residential (apartment complexes and condos) will have similar issues. 100% EV charging will be a cost that these private owners wont pay. And the cost will be just as staggering as the case in 3 above. Imagine a 400 unit apartment complex with 400 100A spaces. Then imagine them all on at 8PM at night. A draw that will certainly exceed the capacity to supply.
[5] The suburban subdivision with 500 homes will have the same problem, the existing electric delivery infrastructure will not handle hundreds of 100A active connections. More millions in infrastructure adds.

While there will always be the rich exception among us, most EV charging will be most economically be done in a shared facility. Some commercial, some shared (HOA run, apartment run, etc) and some done by a governmental authority.

And these facilities will only be economical if the EVs are AVs as well. Using an app the AV EV will schedule itself for a recharge, travel to the recharge site autonomously and return back 'home' (or work or ...). No human, including EV owner, needs to be involved in the process. It just happens.

And these recharge facilities will cost a fraction of the 'build out everywhere' alternative and have very high utilization which will further reduce the end user share of cost.

I just cannot think of any other alternative that can economically work in the mass adoption case.

While yes, within a decade cars will be able to wander off at night to charging depots, it's not quite so bad before that. The low cost and low emissions of electric travel can easily justify a $1,000 cost. Electrical cost for driving at 12 cents/kwh is just 3-4 cents/mile, compared to 7 cents for gasoline in a Prius and 12-20 cents/mile in more typical cars. For the average 12,000 mile/year car that's $1200/year in savings at least in energy cost. (Of course you also paid up front for the battery you will use up which is the larger cost.) A $5K cost is still worth it but takes a lot longer to pay back. (I should not the saving is slightly less if you do a lot of charging outside your house, ie. lots of road trips.)

As noted, for many people, 120v 12a charging actually is adequate but can cause more range anxiety. It can be much cheaper to install a 20a circuit for 16 amp charging -- gaining you around 6-7 miles an hour or around 70 miles per 11 hour night. This can be cheap because you can do a lot yourself, you can buy heavy duty 10/3 extension cords which can run for long distances. Another option is a product like the Dryer Buddy which shares your dryer plug -- the car is switched off while you run the dryer, which you usually don't do in the middle of the night anyway, or for very long.

For the parking lots, you don't need 100a per space. 20a to 30a per space is adequate, and it's actually reasonably cost effective if you are installing 50 of those at once, the cost for each one more reasonable. But yes, it's a lot of power. So I think the future is taxis which worry about going and getting power when they need it.

I don't disagree with your individual points but your reply doesn't address the core EV problems with mass availability:
[1] Urban spaces and many suburban spaces have NO place for a personal charger.
[2] Most families do not have a 'spare' $1,000 to $5,000 for charging even if they have a space.
[3] Governments don't have the funds for mass implementation of public charging in urban areas.
[4] High density EV charging will overwhelm local transmission capacity.

Until mass charging is figured out, EV's will be predominately suburban.

As I said, for those who have no space to put charging at their home, an EV may not be a great choice. But car ownership is much higher among people who do have that space.

Charging at "level 1" (about 5 miles/hour) is actually fairly reasonable for the urban dweller. As I said, the average urban electric car probably drives under 30 miles/day -- on average. If your battery is Tesla sized, even picking up 50-60 miles/night from level 1 (ordinary plug) handles the vast majority of your driving. Superchargers handle the rest.

Now, some people can't even get a level 1 extension cord to where they part their car. If they can't do that, and have no way to charge where they park during the day, then an electric car is probably not for them. But getting in level 1 is a much easier problem than level 2 for homeowners without garages and building parking lots. Not zero problem, but much easier.

For those people, wait for the robotaxis.

If the demand is there, financing instruments will appear to give people the money they need to install charging, paid over time from the gasoline savings. That should not be an issue. Money is not the problem, it's the wiring and parking space. If you don't have a parking space this is not for you.

Governments should not be needed, other than when they run parking lots.

4 is less clear. Again, robocars just drive to the local transformer substation to plug in so you don't need new local charge infrastructure. Because cars can charge only at night and level 2 cars can even constantly turn on and off charging based on electrical availability, you get maximum use of your infrastructure. At night, it is my understand that there is tons of spare capacity with all the air conditioners, lights, dryers, factories and ovens off.

Add new comment