I‘m based in Germany and I‘m trying to build my own charging station from scratch, but didn't want to go witih a "finished" solution like openEVSE, so I learn more about the communication „protocol“.
However, I ran into some problem and can‘t get the car to accept my charger.
First of all, the solution I wanted to achieve is:
- use available solar power for charging, which will otherwise be sent to the grid (I already have my own energy management system for this, so I have the necessary data already available)
- be able to go as low as 6A 230V single phase (1.4kW)
- be able to charge "full speed" at 16A three-phase (11kW, because of dynamic electricity pricing, where the price changes every hour and can get really low)
I read through the openEVSE theory of operation where the J1772 standard is explained, have drawn my own schematics (relevant parts attached as pictures) and wrote my own piece of software on an arduino.
However, I could not get my Tesla Model 3 to charge. The process looks like this:
1. sent constant +12V to CP pin (standby, state A)
2. connect car
3. voltage on CP drops to +9V
4. wait 250ms, then start PWM (state B)
5. It takes about 400ms, then car requests a charge, peak voltage drops to +6V
6. wait 50ms, then close relay and send power to the car (state C)
and here something happens what I don‘t understand. After about 20ms (in some rare cases also 50ms or a bit longer) the car sets it‘s status back to B, then after about 400ms back to C, after 20ms back to B and so on ...
This happens exactly 4 times before the car stays on B and displays an error message to check the charger.
What I already tried is to close the relays immediately when the car changes to C, but although the relay closes and the car displays a voltage for a short amount of time, it still goes back to state B with the error message.
I made some oscilloscope screenshots/photos which are also attached, and for me it looks okay, but maybe I‘m missing something ...
So maybe one of you guys could help me out here, although it‘s not an openEVSE issue?
How is your ground (Earth) setup? You will need Earth from the AC Input, EV output and your controller reference all tied together.
Thanks very much for the reply.
I just checked again to be sure that earth is really connected. L1, L2, L3, Neutral and Earth is coming from the AC side. Earth is always connected to the Earth pin of the Type 2 plug (<0.1 Ohrm). Also the control voltage ground is connected to earth, so the CP signal is +/- 12V referenced to earth.
I'm also struggling with the correct resistor value at the opamp for the CP signal. I found schematics with 620 Ohm, 910 Ohm and 1k. For testing I put resistors parallel to the 1k resistor to match the mentioned values, but still no luck.
Is it important to match the voltages +12V, +9V, +6V, ... exactly? Because with the 1k resistor I'm slightly over +5V when the car requests a charge for the short amount of time, with the 620 Ohm resistor it's around +7V.
Your Operational amplifier may have internal resistance. The total resistance Opamp/resistor should be 1k equivalent.
It is the voltage that is important, except 12V on State A does not matter at all (nothing is connected to know)... 9V and 6V should be as close as possible. Most cars do not support State D Ventilation required so 3v is not so important either but should be within specs if possible.
The SAE J1772 Standard requires the voltages to be 8.36 - 9.56 for State B and 5.48 - 6.49 for State C.
I tried adding parallel resistors of a few kOhm to the 1k resistor, but in the end I just increased the input voltage to about 14,1V and it seems to match the voltages perfectly with the 1k resistor. +8,8V in state B and +6,0V in state C (see attached images).
But still no luck getting the car to charge ...
Maybe tomorrow I will try to capture the CP signal of the original charger, to see if it looks any different.