TMC is an independent, primarily volunteer organization that relies on ad revenue to cover its operating costs. Please consider whitelisting TMC on your ad blocker or making a Paypal contribution here: paypal.me/SupportTMC

Tesla's 10kw Charger

Discussion in 'Model S: Battery & Charging' started by Hogwit, Feb 29, 2016.

  1. Hogwit

    Hogwit Member

    Joined:
    Feb 28, 2016
    Messages:
    9
    Location:
    Clinton, CT
    I have a Coda sedan which, unfortunately, is limited to only the standard 6.6kw J1772 with no quick charging capabilities. The chargers onboard are two 3.3kw Lear chargers (the same as the Chevy Volt) in parallel.
    I was wondering if anyone here has any experience in essentially trying to "hack" the Tesla's onboard charger and if so what would be needed to get one (or two) to work outside of a Tesla.
     
  2. FlasherZ

    FlasherZ Sig Model S + Sig Model X + Model 3 Resv

    Joined:
    Jun 21, 2012
    Messages:
    7,019
    There are several people working on the CAN messaging that takes place, but I don't believe anyone has been working on the messaging that makes the charger work. The unit is not self-contained - it's an intelligent unit that communicates with the car's other systems, so it would take a lot of work.

    It's not impossible but will take a good bit of reverse engineering, and you'll likely have to create a supervisory unit for it.
     
  3. wk057

    wk057 Senior Tinkerer

    Joined:
    Feb 23, 2014
    Messages:
    4,715
    Location:
    Hickory, NC, USA
    Yup, definitely more complicated than the Leaf and Volt chargers.

    I have the one on my bench "working" (I can make it charge a loose Chevy Volt battery pack I have), but it's going to require a lot of work to make it actually usable. Specifically, going to need an external and accurate high voltage sense because the charger relies on the Model S's BMS for voltage cut off points and internal calibration of its own DC voltage readings. It takes the emulation of several other modules in the car on the CAN bus in order to get the charger to agree to work. It also requires J1772 with the corresponding full EVSE (AC contactor(s), pilot signal, etc) and such to function.

    Long story short, it's a bit of a pain and I've kind of back-burnered the project given the need for a bunch of additional work to make a proper control board for it.

     
  4. Hogwit

    Hogwit Member

    Joined:
    Feb 28, 2016
    Messages:
    9
    Location:
    Clinton, CT
    Basically what I was thinking was along the same lines as what EVTV did to the Lear charger...see here: http://store.evtv.me/proddetail.php?prod=Learcharger&cat=23 and there is a YouTube video of it all somewhere, but I can't seem to find it right now.
    What they did was have an Arduino mimic the signals that the car would've given the charger and now they are selling the charger with the Arduino preprogrammed as a complete bundle, I was wondering/hoping that something similar could be done to the Tesla charger, meaning, having an Arduino mimic the car's signals and command the charger(s) to turn on.
    Do you think that something like that could be done, even if it would require an external voltage sensor? What was required to make your charger work, and aside from the fact that it's currently a mess of loose wires, why is it not currently usable
    Also, are you familiar with what signals the Lear chargers need? Do they receive an external voltage reading that could be used in place of a second sensor?

    Mostly unrelated, but any chance you, or someone else, may have a spare charging connector off a Tesla? I'd like to put it on my car for if/when a Tesla HPWC is available but not a J1772 and just use Tesla's adapter for charging at standard EVSEs.
     
  5. wk057

    wk057 Senior Tinkerer

    Joined:
    Feb 23, 2014
    Messages:
    4,715
    Location:
    Hickory, NC, USA
    #5 wk057, Mar 1, 2016
    Last edited: Mar 1, 2016
    I want to mentioned first, and I've had this conversation with Jack... using Arduinos for this stuff, especially beyond the dev stage, probably isn't the greatest idea. These things are really not made for an automotive environment. At best, arduino is more intended as a dev platform to help get ready for a production version of a product, not something I'd ever install in my car... especially not controlling a critical system like a HV charger.

    That said, no, it's going to be way more complicated to get the Tesla charger going properly and safely vs just strapping a CAN-enabled arduino to it and calling it a day. The charger relies on the Model S BMS for quite a bit. HVDC voltage and current readings from the BMS affect the charger. It expects the DC contactors to be opened and closed when the BMS signals they are, otherwise it faults and needs to be reset. It expects communication with the fast charge equipment in the car's HVJB and expects to be able to perform tests on that equipment, which is directly connected to the master charger and NOT via CAN. It expects communication with the charge port and expects the charge door and latch to be in an appropriate state and sequence of events in order to charge and or perform the fast charge equipment testing (ie, it would be a bad idea in a real car to test the fast charge contactors while the charge door was open and someone could electrocute themselves). The charger also needs a DC precharge circuit (normally inside the battery pack) since the DC side capacitors are not switched. There's also the high voltage interlock loop that needs to be faked for the charger to work, including the constant current signal. The charger also expects the BMS to announce the current HVIL current and voltage on CAN, and it expects to be able to perform tests of the HVIL at any time. If the "BMS" tells the wrong current, wont change the current for testing, etc, the whole charger faults and shuts down.

    Additionally, the Tesla charger expects to be connected to a full J1772 EVSE. This is probably fine and actually would save a step when implementing in a vehicle, but is adding complexity to my project since I just want to power it from AC directly.

    Further complicating things, the charger has a coolant flow sensor and expects the CAN data to match the actual coolant flow AND temperatures from the thermal controller in the car to match within a reasonable margin of what the charger is seeing. The charger also can command the coolant pump enabled/disable or increase/decrease coolant flow and it expects to see this change on its own sensors.

    Then there is the high voltage isolation resistance measurements. The charger expects the BMS's HV isolation measurements to match its own. Oh, and the charger won't tell you what its is, so, you have to tell it the actual measured value at any given time within some margin of error, otherwise... fault!

    Lots of this can be done with just CAN and some clever software, but a lot of it needs actual hardware to emulate signals and conditions the charger expects in certain states. It took me quite a lot of work just to get the thing to charge at 120V/5A for a few minutes.

    Then add to that firmware changes. Tesla changes the firmware on these chargers all the time, sometimes drastically. For example, one charger I have expects a -100V offset to the DC target voltage. Another doesn't. Feed the latter a CAN message from the former's data set, and it'll fault and/or overcharge the battery.

    Long story short, strapping an arduino to this thing is not going to make it work. It'd almost be easier to just make my own charger from scratch.
     
  6. Hogwit

    Hogwit Member

    Joined:
    Feb 28, 2016
    Messages:
    9
    Location:
    Clinton, CT
    I fully understand why an Arduino isn't truly appropriate, for long term use, especially in an automotive environment, and do agree in part that it's not entirely appropriate for commercial use. That being said, and the potential ramifications being considered, I, personally, wouldn't object to an Arduino controlled charger for my vehicle.

    Also, in no way do I doubt the complexity of Tesla's charger, but I take a lot to "accept defeat." Given that superchargers use the (physically) same charger as the Model S, just multiplied by 12, do you know how Tesla implements them? Does the supercharger's individual chargers have different software loaded on them or do they spoof a car expecting all 12?

    Also, any chance you, or someone you know has a spare charging connector that I could buy?
     
  7. wk057

    wk057 Senior Tinkerer

    Joined:
    Feb 23, 2014
    Messages:
    4,715
    Location:
    Hickory, NC, USA
    I do not have any spares, no.

    The superchargers use completely different firmware on these units, which would solve basically all of the problems. Unfortunately I don't have that firmware, nor a good way to even start to write custom firmware. :(
     
  8. Hogwit

    Hogwit Member

    Joined:
    Feb 28, 2016
    Messages:
    9
    Location:
    Clinton, CT
    It seems to me that one of us needs to befriend someone either installing a supercharger station or a dealer and get them to reflash some chargers for us.
     

Share This Page