Saturday, January 1, 2011

J1772-2009 Charging for Your EV



This is our last show. Actually our last show for 2010. We'll meet this weekend to decide what to do about 2011.

In this episode we continue the fascinating work of building battery boxes for the Speedster. While not an exciting topic, in converting almost any car to electric drive, most new builders are surprised to learn that making the car run on battery power is almost trivial. But 50% of the work centers on placing batteries and building structures to hold them.

The other 50%? Almost entirely environmental. Heat. Cooling. Lighting.

Mounting a controller and motor and getting it to run - you can nearly enough phone it in.






So to avoid making this episode a total yawnfest, we've decided to tackle a probably useful element we promised almost a year ago when we did our episode on the new J1772-2009 standard/plugs/connectors. That is, how to wire your car to actually do J1772 charging.

With the Chevy Volt and the Nissan Leaf finally shipping....er.... more or less shipping....we think you'll finally begin to see some J1772 charging stations begin to appear.

We build our cars to charge quite automatically from either 120VAC single phase or 240vac two phase power. And we almost always use a NEMA 5-15 recessed male connector on the car so that we can universally connect the car anywhere and to virtually any ordinary household extension cable - albeit you do generally want to use a heavy one.

This is mildly illegal in most respects. First, NEMA5-15 is entirely meant for 120vac operation at 15 amperes or less. In practice, we charge the Speedsters at about 25 amperes and 240vac. We simply make up a cord or "adapter" with a NEMA L6-30 or NEMA 14-50 connector on one end, and the usual NEMA 5-15 female on the other.

This is not good frankly. But it is very convenient. If you go to gramma's house, you can still plug in and get a toke on her good juice, even though there's no charge station there. We buy the best cables and connectors we can find, and while they get warm, so far no incidents.

But it isn't really kosher. And soon it will even be a disadvantage. As J1772 charge stations start to appear, at the least in the garages of friends with Nissan Leaf or Chevy Volts, and potentially public charge stations as well, we would be left out of the party.

For a "standard" I am perplexed by how "closed" all this is. The greed is palpable as the charge station vendors strive to "protect" their ground so they get prices as high as $5000 for what is little more than an AC receptacle. The "standard" does provide some safety features to keep the cord dead until it is connected (which I put in our Texaco Fire Chief charge station nearly two years ago). And the plug and connector ARE kind of quality. But it is amazing how difficult it is to find the basic connector hardware - nearly a year after adoption.

Indeed, SAE itself has been piggishly greedy - you can't find an SAE J1772-2009 PDF online anywhere. They are "selling it" on their web site with digital rights management and dire warnings about copyright infringement. Copyright infringement on a public standards body? This is just inexcusable.

In any event, there IS a conspiracy to control these charging stations and maintain high profit margins on the charge stations. The connectors are a couple of hundred dollars each depending on how long and what kind of cable. None of it warrants $5000 or even $2500, I don't care how many approvals it carries. And the whole concept of having a licensed electrician and permit to install it is laughable. Some of them actually just plug into an existing NEMA receptacle just as we do with our cords. In any event, adding a 60amp or 100 amp circuit breaker to your panel, and running a piece of AWG10 orange 3-wire plus ground to a convenient place in your garage, is work ANYONE can do AND SAFELY.

There are numerous home wiring books at the local hardware store, the wire, the NEMA receptacles, and the circuit breakers. And in almost all localities you are entirely legal to do such wiring ON YOUR OWN RESIDENCE without license or permit.

But back to the car. If you buy a J1772-2009 inlet connector and mount it on your car, and connect the two phase pins and the ground pin to your charger, you won't ever be able to charge your car at any of these charging stations.

This is because the car has to provide some "signals" to the Electric Vehicle Support Equipment (EVSE) aka Charge Station to turn it on. These mysterious "signals" are actually trivial They are actually a single diode, two resistors, and a switch. Here's the diagram:


Click on the diagram to blow it up.

For those who find wiring two resistors, a diode, and a switch, a challenge, you can buy ready made adapters from Rush Dougherty at http://www.TucsonEV.com. He has the connectors you'll need anyway, including a nice plug with a full 8 meter (25 foot) long cord. And he has made a little adapter box with switch and light as well.

With our wiring, or Rush's box, you simply connect the J1772 plug to your car, and then turn on the switch. This drops the sense voltage to 6 volts in the EVSE and triggers the unit to turn on the 240vac power.

When you're done charging, you hit the switch to shut OFF the power, THEN remove the cord. Very much a safety improvement over our current methods.

And it will allow you to configure YOUR electric vehicle to use the new charge stations wherever you find them.

In a future video, we'll go into some detail about how you can set up a J1772 charging station of your own, for HUNDREDS of dollars instead of thousands. We're just going to add a cord to our Texaco pump, and switch on the power manually after you connect the cord as we do now.

But another viewer, David Kerzel of Modular EV Power LLC is working on a printed circuit board control circuit for a home charge station with all the bells and whistles of J1772-2009. We'll follow his progress and hopefully have something to show in a future episode for you to build your own charge station at much less cost. His design will use all UL listed components in the contactor and connectors and so forth, and he intends to include instructions on how to get it permitted and inspected and approved if that be your hearts desire.

See you next year.

Jack Rickard

23 comments:

  1. Hi Jack, Brian & Matt! Happy New year to you all!

    Jack, Do you need the switch for the J1772? The sence/prox. pins are shorter so they make last and break first allowing it to kill power before drawing an arc.....

    Cheers

    Simon

    ReplyDelete
  2. Happy New Year - thank you for sharing your experiences and insights so freely. Keep up the good work.

    ReplyDelete
  3. The official title of the document is SAE J1772-JAN2010. It supersedes SAE J1772-NOV2001. The way J1772 works is first make last break is the ground pin. The last pin to make first to break is the control pilot.

    When the plug makes it all of the way in the latch switch closes completing the proximity circuit. This signals the car a charge is about to start. The car disables the drive and the charger closes the charge request switch telling the EVSE to turn on.

    Pressing the latch opens the proximity circuit signaling the car that the cable is about to be disconnected. Within 100mS the car is to open the charge request causing the EVSE to shut off. All before the plug has a chance to move.

    ReplyDelete
  4. Randy C

    There is no latch switch on a J1772 plug, the pilot and prox. pins are the last to make and are the first to break....

    There is only 5 pins/socket inside the plug and socket, I have a set!





    i

    ReplyDelete
  5. Hi Jack and Brian,

    I noticed you are not strapping batteries together in front box or belly pan. Only a problem related to overcharging?

    Also you mentioned 51 batteries and 151V = 3.35V/cell. Is this what you charge CALB's to?

    Thanks for your great shows.
    Mark

    ReplyDelete
  6. Sorry, the above should read 171V not 151

    ReplyDelete
  7. Simon:

    The point of all this as nothing to do with drawing an arc. In fact, my Texaco Fire Chief charging station is set up this way. Both phases are on contactors that are normally OFF leaving the cord entirely dead. I plug the cord into the car, and flip a switch on the side of the gasoline pump. This applies 12v to the coils of two contactors, which close, applying the two 120vac phases through the cable to the car.

    That's the point with J1772 as well. The cable and plug are DEAD except for the 12v. Plug it into the car, and hit the switch (in this case on the car). The contactors again close and apply the 240vac to the car. Hit the switch again, and they open. THEN you can remove the plug.

    It's a good idea all the way around.

    Technially, you COULD reduce this wiring even further to a diode and an 882 ohm resistor, and it would indeed trigger most J1772 plugs to charge as soon as you connected the plug.

    Jack Rickard

    ReplyDelete
  8. Randy C:

    Our car really doesn't need or want to know all that, which is why we didn't deal with the contact switch on pin 5. It doesn't do anything with the EVSE, it is there to signal the car. We have no circuitry to detect it, and frankly don't need it.

    Perhaps one day will do a feature on implementing this to disable your controller so the car can't be driven off with the cord connected - again not a bad idea.

    Our purpose here was a quick and dirty circuit that would let you actually plug into an existing charge station and signal it to charge your car. We don't, for example, key our charger to limit current to what's available on the PWM, under the assumption that any charge station you plug into will offer more current than your charger needs.

    Jack Rickard

    ReplyDelete
  9. Mark:

    No, we no longer strap batteries at all. If your cells are swelling, they are telling on you. You have overcharged them. They do not swell from overdischarge, only from overcharge. We've rather ceased overcharging them and life is better this way.

    We charge CALB cells to 3.47 or 3.50 volts per cell. So 51 cells will be charged to 178.5 volts. As soon as we terminate the charge process, they invariably fall back to about 3.35 v per cell or 171 volts.

    The open circuit voltage of a FULLY charged cell is 3.40v in all cases.

    Jack Rickard

    ReplyDelete
  10. Being very novice could you address charging under wet conditions, such as a charger in a parking lot in the rain and a wet car in a garage.

    Are GFI circuits used at the charging stations?

    ReplyDelete
  11. The charge stations do have GFI in them and indeed you can get GFI circuit breakers for your box for that matter. There is no magic there.

    As to charging in wet conditions, this is what a good bitof this is about. It really shouldn't matter. BUt in rain and snow, if you picture this equipment as old, worn, and cracked and frayed, there is a potential issue. That's what the "dead cable" is about.

    It is not like a circuit in your house where one leg has power and you simply haven't thrown the switch to complete the circuit. Both phases are physically disconnected from the cable. There is no power in it. You plug it in, and using this switch, switch on the power. THEN the cable becomes live, but you're not touching it at that point.

    The physical design of the plug rather keeps the fingers clear of the contacts in any event. But the device is safer at little real cost or inconvenience.

    Jack Rickard

    ReplyDelete
  12. I don't see why the boutique store is so attractive for marketing a car. You stated in this weeks video that years ago you wondered why the car couldn't just be configured on line and delivered. If 3000 people have preordered a vehicle that isn't even made yet, how would having hundreds of stores and staff help get a car to market.

    Of course they are going to need some service facilities but third floor of the mall isn't where I want to get my staticy radio serviced.

    A fun android app to keep people informed and excited would cost less than the heating bill in the Zürich store.

    No show room, just a service bay and waiting room/loaner car lot, good to go.

    By the way work on some better lighting for the video of the work on the car.

    ReplyDelete
  13. I think a soft box would solve your lighting issues. Perhaps suspend it just above the camera so there's no stand to trip over. If it's on a pulley you can raise it to keep it out of the way during non-shooting hours.

    http://en.wikipedia.org/wiki/Soft_box

    ReplyDelete
  14. Jack said:
    "No, we no longer strap batteries at all. If your cells are swelling, they are telling on you. You have overcharged them. They do not swell from overdischarge, only from overcharge. We've rather ceased overcharging them and life is better this way."

    I know that BMS is a bad word to you Jack but this comment got me thinking. Put some pressure sensors on the batteries. Might as well add heat smoke detectors also. Wire them up to cut all power and call 911 at the first sign of trouble. That might keep some cars from burning down.

    Brad Lindberg
    Overly Critical MN Viewer

    ReplyDelete
  15. Heat detectors sound like a nice idea: unlike a BMS, they would have failure modes that didn't result in a fireball or dead cells. Still 50 or a 100 wires and sensors to break, and 100-200 connectors to corrode....


    Another good vid this week guys

    ReplyDelete
  16. Hi Jack.

    I really enjoy the show each week. I've watched them all in the archive now.

    I noticed in this weeks show, and I think lasts, that you had the camera on auto focus. in your new place with the car in the background the camera was periodically shifting to focussing on the car and not you guys. You should probably set and lock off the focus.

    Any news on the Smart's hub motors? I'm looking forward to that experiment. I like the idea of hub motors. I know lots of people talk about unsprung mass and that they may not be reliable, but I really like the idea.

    Volvo were messing with them, but they seem to have dumped the idea, so I guess maybe they don't work in the real world. Still though I'm interested to see how you get on with them.

    Nick F. :-)

    ReplyDelete
  17. I just came across another instance of a fire "of undetermined cause" starting at 3:00am - saabrina right here on blogspot. Anyone who has watched even one of Jacks videos knows what that means and lo and behold we see AGM batteries with BMS.
    Surely it must be a trivial effort to install thermal cutouts to the charger. I note that the BRUSA is actually supplied with thermisters and the zivan has a temperature probe too. Really I suppose it would only be necessary to monitor a few locations within the pack to achieve a certain level of safety. Smoke detectors, either in the car or at least the garage, would at least give some warning too.
    Supposedly we will never be in this situation with an undercharging regime but sooner or later something is bound to go wrong, so -safety first.

    ReplyDelete
  18. Yes belt-and-braces sounds a good idea.

    ReplyDelete
  19. Jack many thanks, you are doing a great work for the world.

    On your circuit, I believe you said the duty cycle of the PWM tells you how much current the charger can deliver. What if you plug into a charger that can only deliver small Amps, so it has a small duty cycle on the PWM? It seems your resistor network might then make the average voltage too low. Now suppose you go to another charger with 100% duty cycle, now your resistor network will report too high an average voltage and the charger will get too high a voltage and not start.

    If this is the case, you could fix this by doing a bridge rectifier instead of a single diode (provided the signal is either +12V or -12V and doesn't hang out near 0).

    ReplyDelete
  20. A great Blog, thanks...

    We are a UK Charity donating charge points to public locations. We have recently started a project ("OpenChargePoint") to develop a ultra low cost J1772 EVSE. We would be interested in working with like minded people on this project. I've reached out to David Kerzel, and would be happy to hear from anyone else who's interested.

    ReplyDelete
  21. I recently got my hands on the IEC 61851 controller of the type being fitted to the european 3 phase charging points. It uses a very similar system to J1772 with a few small differences. First , the Proximity pin (PP) must connect to earth via a diode. I used an 1N4004. This brings up 12v dc on the communications pin (CP). To enable charging this must then be pulled to 6v. I used another 1N4004 and a 1k 0.25w resistor. No need for a switch in this case as it does not need to go to 9v first. Might be of interest to people on this side of the water. Its a 7 pin plug with 3 phase , neutral and earth and the two pins outlined above.

    ReplyDelete
  22. The duty cycle of the PWM would be accounted for on the EVSE side, but I don't think would matter in any event.

    I gave some thought to the current intelligence and for the most part decided it didn't matter. This is kind of an advanced topic, in that it requires your charger to tailor it's charge actiivity to not exceed the available current.

    For the most part, the minimum J1772 current capability seems to be 30 amps, and we don't have any cars that charge past 24 amps anyway. Most of the J1772 we think will be done at the maximum of 80 amps. And so there really isn't much need to current limit to accommodate the EVSE at all. Your charger would have to be at least 7200 watts to trip a 30amp breaker. And at 80 amps we're talking 19200 watts - my dreams to come true! Precisely the problem I would LOVE to be dealing with.

    Unfortunately, our charges tend to be 3000 or 3600 watts, half of what would be necessary to strain even a 30 amp breaker. We do have one PFC-75, and of course it can throttle current to whatever you like, but it is very manual.

    So to charge using J1772, we can for the moment ignore the PWM entirely. I can only DREAM that it might become an issue in the future.

    Jack Rickard

    ReplyDelete
  23. Maybe too late to tell, but ... maybe there is an error in the diagram drawing on this page. I believe that should not be considered:

    R1 / / R2 = 877 Ohms

    But instead:

    R2 / / R3 = 877 Ohms

    best regards!

    ReplyDelete