Trouble with control pilot signal (i think..)

[Lars-Erik Eriksson]

Is the negative part of the 12v supply important for the charging function?
As Nissan Leaf's original charger don’t use the -12v (at least not at to 9V stage, where I could measure it), I skipped it on my DIY EVSE too.
My EVSE makes the connection sequences correct, and energizes the mains and sends a nice 1kHz 6v 30% duty cycle signal on the control pilot .
But my Leaf wont charge. It beeps once at connection, as it use to, but then immediately beeps twice and doesn’t charge.
Could this have something to do with the missing -12V?
If you have any idea what to do ill be grateful.
Maybe you don’t want to help me as I have built my EVSE on an Atmel Mega16. But I am used to the M16, and have managed it to adapt the duty cycle to actual sunshine intense (I charge my Leaf with my solar panels).
I have the 1k in series with the control pilot and get all the sequences. But now I have got stuck on this.

[chris1howell .]

It sounds to me like the pilot may be okay. -12v is used for the diode check, it is important but is a commonly excluded.

Does your j1772 cable have the correct resistors for proximity? On the leaf both proximity and pilot have to check out before charging will start.

[chris1howell .]

Just for clarification your DIY EVSE should be sending a 12v pilot signal to the Leaf and the Leaf should be dragging it down to 9v then 6v for charging.

[Lars-Erik Eriksson]

Thank you for answering. Yes, the cord has the right resistors (150 ohm to ground while not pressing the button, 480 ohm while pressing the button). When i measure the resistance in the original Nissan cord i get the same values.
When i test my EVSE on my bench, and manualy lower the control pilot to 9v (via a variable resistor to ground), the control pilot switches from stable 12v to 1kHz sqare wave. When i lower the control pilot to 6v, the mains relay switches on.

Do you know if there should be a notiable delay in this sequence while connecting the car (from connection of the cord to the car to the mains contactor switching on)? Now the mains contactor switches on imideatly as i attach the cord to the car, so i dont have a chance to measure the 9v.

Grateful for answer.

[chris1howell .]

I would not use variable resistors, set up a static resistance of 2.74k for State B connected so you see how close to 9v you really are. For Charging State C use 882 and measure the actual voltage. Before I had the correct values, I found 5 or 6 resistors in my junk box and put them in parallel and series until I got 2.74 and 882, kinda ugly but it worked. Depending on the OpAmp you used you may need to adjust your 1k accordingly. OpenEVSE uses a LF353 opamp which is not rail to rail so to adjust for internal losses we use a 649 ohm pilot resistor to have the correct voltage at each state. Your Opamp needs to have a high slew rate so the 1khz square wave is correct a slow slew rate will give you a sawtooth wave.

Timing and the correct handling of transition states is very important, some vehicles such as the volt are very sensitive. The Tesla on the other hand is very sensitive to pilot voltage. You can find the required timing in the J1772 document. You may want to try about 100ms delay before activating the relay.

--
You received this message because you are subscribed to the Google Groups "OpenEVSE" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openevse+u...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[Lars-Erik Eriksson]

Thank you Chris for sharing your time, knowledge and experience!

I really appreciate that.

Now I have homework to do!

 

As I don’t use the -12v (I begin to regret that) I use a Darlington driver (uA9668) instead of an opamp. The uA9668 has good "rail to ground".

But I sure will check the voltages as you describe, and if needed adjust the 1k resistor.

It looks to me as the slew rate is good enough, but what do you think (see Picture)

Thanks again for your good property to help strangers.

Den fredagen den 27:e december 2013 kl. 19:33:07 UTC+1 skrev Chris:

I would not use variable resistors, set up a static resistance of 2.74k for State B connected so you see how close to 9v you really are. For Charging State C use 882 and measure the actual voltage. Before I had the correct values, I found 5 or 6 resistors in my junk box and put them in parallel and series until I got 2.74 and 882, kinda ugly but it worked. Depending on the OpAmp you used you may need to adjust your 1k accordingly. OpenEVSE uses a LF353 opamp which is not rail to rail so to adjust for internal losses we use a 649 ohm pilot resistor to have the correct voltage at each state. Your Opamp needs to have a high slew rate so the 1khz square wave is correct a slow slew rate will give you a sawtooth wave.

Timing and the correct handling of transition states is very important, some vehicles such as the volt are very sensitive. The Tesla on the other hand is very sensitive to pilot voltage. You can find the required  . You may want to try about 100ms delay before activating the relay.

[chris1howell .]

It is my pleasure to help and the reason OpenEVSE was started...

The slew rate looks fine. I tested early OpenEVSE builds with a very slow opamp and it worked fine but the duty cycle was a bit off. I think all that is left is to check the voltage levels.

[Lars-Erik Eriksson]

Homework done! Works perfectly!
You were quite right, it was the voltage levels.
I had to adjust the 1k and change my detection of the control pilot.
You are a hero in your unselfishness!
Millions of thanks!