OpenEVSE - Open Source Charging Station

[chris1howell]

Ford Focus EV just added to the OpenEVSE compatibility list... Thanks for the report David.

List now includes:

Leaf
Volt
i-MiEV
Plug in Prius
Focus EV

 

[waidy]

Ford Focus EV just added to the OpenEVSE compatibility list... Thanks for the report David.

List now includes:

Leaf
Volt
i-MiEV
Plug in Prius
Focus EV

What about Tesla Model S?

 

[davewill]

What about Tesla Model S?

I'm sure Chris will test his ASAP.

 

[mitch672]

Model S comes with a J-1772 adapter to the Tesla HPC connector, I'm sure it's no problem.
What would be interesting is if someone with a Signature Edition Model S with the dual 10KW onboard chargers could test on a 75A OpenEVSE, and see if that works. Teslas proprietary HPC EVSE can use up to a 100A breaker (to supply 90A) to those 2 onboard chargers...

 

[chris1howell]

Model S comes with a J-1772 adapter to the Tesla HPC connector, I'm sure it's no problem.
What would be interesting is if someone with a Signature Edition Model S with the dual 10KW onboard chargers could test on a 75A OpenEVSE, and see if that works. Teslas proprietary HPC EVSE can use up to a 100A breaker (to supply 90A) to those 2 onboard chargers...

Mine is reservation P1089 so it is still a few months away... I already have the 75A ITT J1772 cable and a Monster 75A Square D contactor.

 

[mitch672]

Model S comes with a J-1772 adapter to the Tesla HPC connector, I'm sure it's no problem.
What would be interesting is if someone with a Signature Edition Model S with the dual 10KW onboard chargers could test on a 75A OpenEVSE, and see if that works. Teslas proprietary HPC EVSE can use up to a 100A breaker (to supply 90A) to those 2 onboard chargers...

Mine is reservation P1089 so it is still a few months away... I already have the 75A ITT J1772 cable and a Monster 75A Square D contactor.

Nice Chris! I was thinking about reserving one, 12-18 months out for delivery is what they are quoting now.
Just trying to figure out the best way to get the $82.3K + $5143 in MA sales tax (%6.25 ouch). was looking at the Standard Model S 85KW pack with most of the options. since you are out in CA, you'll probably run into a few Model S owners, maybe you can get one to let you test your "Super" OpenEVSE Don't think it's worth $1,200 for the Tesla HPC EVSE, just to get 8 more amps (dual 20KW chargers could pull 83.3A @ 240, assuming you have a 100A breaker feeding it)

 

[Barbouri]

Cool. I just sent out the new board design today.

Hi Chris. Not sure how to comment on the Open EVSE project page but wanted to make you aware that the DC-DC on the Rev 2 schematic is listed as +5V input but the V2 BOM says +12V

PDF and Eagle schematic corrected. It is a 12vdc input to isolated +/- 12vdc DC-DC converter.

Greg C.

 

[Nekota]

You can shift into neutral with the J1772 connector attached?

Here is the end state for 'ready to roll' in Neutral while plugged in...

 

[KevinSharpe]

Ford Focus EV just added to the OpenEVSE compatibility list... Thanks for the report David.

List now includes:

Leaf
Volt
i-MiEV
Plug in Prius
Focus EV

Fantastic! Any progress with the Tesla 'issues' that have been reported?

 

[KevinSharpe]

PDF and Eagle schematic corrected. It is a 12vdc input to isolated +/- 12vdc DC-DC converter.

Great! Sorry, but I did spot one other item that I forgot to mention... text says "5V Relay" but they are clearly powered by the 12V rail... I know this is all trivial stuff but hopefully useful to someone

 

[Barbouri]

Finished testing the new Open EVSE DIY V3.0 board this weekend.
New in this version is a high efficiency 5 volt switching regulator, charging LED header, additional decoupling capacitors for IC's,
additional filtering for +/- 12 DC-DC converter, and onboard L1/L2 sense circuit to support V 0.5.0 firmware.
Removed in this version is the 6 pin SPI header, digital 3,4,+5 connector, and 7805 linear regulator.
12 volt idle current with Adafruit RGB LCD shield on is 48 ma (0.58 watts), and charging with relay energized is 195 ma. (2.4 watts)
I still have a lot of documentation to finish, but so far very happy with the V3 DIY board.

Open EVSE DIY V3 board

Open EVSE DIY boards V1, L1/L2 sense, V2, V3

Greg C.

 

[eaf]

Hi,

I have a few question regarding GFCI schematics... I apologize if they sound dumb, but please explain this to me... I'll use v2 for notations, but the schematics is the same in v6 except that the opamp isn't labeled.

It seems that it's an inverting amplifier/level holder followed up by a comparator, both provided on a single LM358 opamp running from a single supply. When CT senses current, it generates a voltage spike of 330*I/1000 on R17, which is limited by zeners and subsequently amplified with gain 100=R16/R20. In order to trigger comparator C9=10uF will have to gain at least one volt, and given max output current of LM358 30mA, this will take at least 0.3ms. The reaction time is good. But the current seems a little high. 1V on C9 requires 0.01V on R17, i.e. 30uA through R17, which, assuming 1:1000 CT (the one that's used in the examples) means 30mA GFCI trigger current. Do you think that's a little too high? Class A GFCI should trip at 5mA...

Second question is about the amplifier. LM358 here is powered by a single power supply (as opposed to split). And in equilibrium both + and - are held at the GND level. When CT triggers, - seems to be driven below GND whereas the absolute minimum voltage for either - or + from LM358 specs is -0.3V. Should +/- not be held at VCC/2 level for single power supply use? Do I understand it correctly that - will indeed be driven below GND when CT is triggered?

Third question is regarding GFCI circuit output. It is fed into ATMEL where it triggers an interrupt, and then the software turns off relay and does some pilot signaling. Would it be more reliable not to depend on software to turn off power to the car in case of GFCI faults? E.g. perhaps have a flip-flop after GFCI circuit, which output be subsequently ANDed with PD6? This way no matter what state ATMEL is in and what software is doing, the AND would immediately turn the relay off?

I like this project a lot, just want to understand it better.

Thanks!

 

[eaf]

I noticed that in the through-the-hole layout there is a JFET LF353 instead of LM358, and thought at first that "-" at GND or below would work, but then found this in the datasheet: "However, neither of the input voltages should be allowed to exceed the negative supply as this will cause large currents to flow which can result in a destroyed unit"

Also I have one question about comparator... When comparators are made of opamps they usually take signal on "-" and have a positive feedback, which works both as a hysteresis circuit and as a switch accelerator: when "-" is raising slightly and slowly, opamp starts moving its output downward, which adds positive feedback to "+" and quickly forces comparator to low output. There is no much need for a hysteresis here, but how about slow raising fronts? Will ATMEL detect such fronts and raise interrupt?

 

[chris1howell]

Hi,

I have a few question regarding GFCI schematics... I apologize if they sound dumb, but please explain this to me... I'll use v2 for notations, but the schematics is the same in v6 except that the opamp isn't labeled.

It seems that it's an inverting amplifier/level holder followed up by a comparator, both provided on a single LM358 opamp running from a single supply. When CT senses current, it generates a voltage spike of 330*I/1000 on R17, which is limited by zeners and subsequently amplified with gain 100=R16/R20. In order to trigger comparator C9=10uF will have to gain at least one volt, and given max output current of LM358 30mA, this will take at least 0.3ms. The reaction time is good. But the current seems a little high. 1V on C9 requires 0.01V on R17, i.e. 30uA through R17, which, assuming 1:1000 CT (the one that's used in the examples) means 30mA GFCI trigger current. Do you think that's a little too high? Class A GFCI should trip at 5mA...

Second question is about the amplifier. LM358 here is powered by a single power supply (as opposed to split). And in equilibrium both + and - are held at the GND level. When CT triggers, - seems to be driven below GND whereas the absolute minimum voltage for either - or + from LM358 specs is -0.3V. Should +/- not be held at VCC/2 level for single power supply use? Do I understand it correctly that - will indeed be driven below GND when CT is triggered?

Third question is regarding GFCI circuit output. It is fed into ATMEL where it triggers an interrupt, and then the software turns off relay and does some pilot signaling. Would it be more reliable not to depend on software to turn off power to the car in case of GFCI faults? E.g. perhaps have a flip-flop after GFCI circuit, which output be subsequently ANDed with PD6? This way no matter what state ATMEL is in and what software is doing, the AND would immediately turn the relay off?

I like this project a lot, just want to understand it better.

Thanks!

Hi eaf... Your questions are not dumb at all... EV Charging stations have a different trip point than the standard 5ma trip for a Class A GFCI. The requirement for EVSEs is 20ma. The burden resistor (R17) for the CT is set to trip at the circuit at 20ma. C7 and C14 slow the reaction time just enough to prevent high frequency noise from creating false trips. The 10uf C14 may not be necessary at all, removing it will improve the reaction time however testing will be necessary to ensure false trips are not a problem. More background on the GFCI circuit can be found here: http://www.crmagnetics.com/Products...tifier Circuit for CT Signal Conditioning.pdf

You are correct the GFCI requires the processor to function. Actually, the first 3 versions of the boards I built (OpenEVSE started with v4) had hardware logic (2xNOR, 1xAND and 1x XOR) for Fault, Reset and Error output for the Microprocessor. The hardware logic was removed to simplify the EVSE as the ATMEL proved to be extremely reliable. In fact there has been only a single reported issue with hardware failure and it was caused by a bad voltage regulator (sorry GlennD)...

 

[eaf]

Thanks for the link to GFCI schematics. Do notice however that they don't ground "+" in there. Both "+" and "-" float, and even if one of the CT's ends is meant to be grounded, they may imply that a dual PS is to be used (and then the about-ground signal will be right in the middle of opamp operational range, i.e. between VCC- and VCC+). In your case "+" is grounded, i.e. it's already at the low VCC- rail, and for the GFCI circuit to trigger, "-" has to go even lower. Do I miss smth here?

 

[chris1howell]

Do I miss smth here?

I doubt it... but I may have.

 

[electronit]

Hi Chris,

So I'm succesfully using my portable EVSE on our cross-europe journey - thanks ! http://www.grasseglasgow2012.eu

One thing I'd like to add is the implementation of the proximity pilot signal to a) check for connector present in my 62196-2 socket (the proximity pin is shortest - first one to get disconnected - before pilot), and b) limit the amps to what the cable can handle according to the table on:
http://en.wikipedia.org/wiki/SAE_J1772#Signaling" onclick="window.open(this.href);return false;

Now that should be pretty simple, a dividing resistor between +5, analogIn 3, and the proximity pilot pin in the socket should do it. However, since I'm not mighty familliar with the arduino hardware, I wouldn't mind your advice. What voltage readings does the analog input pin give ? eg if I put a 1K5 resistor to the proximity pin and an analogIn, I'd get 2.5V for 16A. What value does an "analogRead" give for that ? Also - I'd be pulling max 3mA from +5V in that case - I'm assuming that should be OK.

Thanks,
Mark

 

[chris1howell]

Mark. Thanks for sharing the link. I am glad OpenEVSE is part of your adventure.

The Analog pins read from 0 - 5V and return a 10 bit value (0 - 1024), so your example of 2.5v would be a reading of 512. As long as you keep the current below 20ma you should be okay. So just pick a resistor that will keep you within that voltage range. You will also need to modify the OpenEVSE firmware to do an AnalogRead and set the pilot. The latest experimental firmware is here: https://github.com/lincomatic/open_evse Feel free to branch the code, it probibly would be good to keep track of Euorpe specific features such as the 62196-2 proximity pilot and changes to the Advanced PS code for 230V.

 

[TEEEKOWED]

I started a Google Code Open Source project for the "Open EVSE" based on the DIY EVSE http://www.mynissanleaf.com/viewtopic.php?f=26&t=2736
I just started populating the site with info and a basic schematic. There is much more to come...

Take a look and contribute if you can.

http://code.google.com/p/open-evse/

:twisted:

IF ANYONE WOULD LIKE INFORMATION RE-EV-CHARGE AMERICA

I WORKED THERE [email protected] 702-782-5064 ANYTIME

CURRENT INFORMATION AND FACTS

 

[mwalsh]

color=#FF0000] :twisted: [/color]

What are you trying to stir up here, Mr. Haynie? Suggest you explain yourself, publicly.