Chademo Controller Development

[arber333]

Tnx yes it seems i misread the schematic as ot is from the back of the connector. I will reset cables AGAIN.

Well i still have to try what it is that vcu wont start can bus.

[CCSknowitall]

You should see 108 and 109 show up about 2 seconds or less after start1/D1 is set. It varies a bit between manufacturers but in general if you don’t see 108/109 after a few seconds, you have your CAN wires swapped or otherwise messed up.

[Bigpie]

Pretty sure can lines are the right way round. Been chatting to damien on his discord, looks like I need 12v on proximity. Will try tomorrow and report back

[Bigpie]

So that wasn't the issue.

Still no EVSE side can messages received. I've double checked my can H and L orientation, as I'm looking at the socket on the car, I have CAN H as the left most pin in the lower group for 4 and CAN L as the right most pin in the lower group of 4. If I connect a can device directly to the socket on these pins, my controller receives the messages.

[arber333]

It just occured to me; maybe i am looking at the wrong source! I read on the forum some Damiens PCBs had bad chips on them. Maybe CAN transciever is bad. I will just snap some CAN test firmware on VCU to verify i got good data off both CAN ports...BRB

[arber333]

Pretty sure can lines are the right way round. Been chatting to damien on his discord, looks like I need 12v on proximity. Will try tomorrow and report back

Why you need 12V on proximity? It is designed to activate off GND signal. Do you mean a pullup resistor? You can draw 12V off a supply pin inside VCU.
EDIT: i consider Pin 7 proximity detect. I use it to power up my VCU and ISA as well. 10K pullup shouldnt hurt there i think.

[Bigpie]

The charger side grounds the proximity, damien uses it via a relay to power up the controller, so wondered if I needed 12v on there. My can logging gets my messages but not the chargers can messages. If I replay the charger messages feom someone's recording into my socket, with the 12v on pin 2, then ground 10, my controller thinks its charging and closes contactor. I'd think I've got can high and low mixed up, bit checked multiple times with multiple sources.

[Bigpie]

Success. Changed the code to set the extended flag to false and away it went. The most expensive half a kwh I've ever purchased.

[whereswally606]

like it BigPie. you got chademo working just in time for the massive price hikes. Still nice to have less charger anxiety out in the wild. Personally id blot out your reg since there are some nobs cloning ev plates in london of late if you watch johnathan porterfield's channel.

[arber333]

Success. Changed the code to set the extended flag to false and away it went. The most expensive half a kwh I've ever purchased.

Can you specify where did you correct this flag? I cant find any reference to extended ID in existing code. Did you input your own flag call?

tnx

[Bigpie]

I just added outFrame.extended = false; I originally forked your repo, I'd have done a PR but my editor changed a load of line endings so it's not showing just this one change.

Have you checked your can bug log to see if yours is the same issue?

[arber333]

Have you checked your can bug log to see if yours is the same issue?

Again i got error in communication.
And also now i couldnt hear contactors click in, though i did hear the small contact relay.

I will need to test this with EVSE simulation.

[Bigpie]

Did you get a can log?

[arber333]

Did you get a can log?

OK my CAN log is attached...

I finally managed to get the EVSE charging.
1. Of course i needed to add simple ID flag
2. I managed to solder 1K 12V pullup to wrong VCU pin I had to resolder it. Now i just pull 12V from input via 1K resistor to IN2 pin (large connector pin no. 6).
Of course now that i corrected both mistakes EVSE pulls contactors in and starts to charge, YAY!

NO... after 24A ramp EVSE throws a flag. It says current request mismatch. When i connected to USB i got 20A provided for -20A VCU request.
I tried to charge several times and later i tried to set reverse current inside the code, but i dont know how...

@Bigpie how is your ISA shunt connected in reference to battery voltage?
@Isaac96 can you help me with the code? How would i reverse the ISA (or VCU) CAN reporting so EVSE would see positive current request?

tnx

[larsrengersen]

Do you have a dedicated current sensor for the CHAdeMO lines? Or an overall sensor for measuring the current going into your pack?
In case of the second option: are there any loads consuming power during charging?
If for example a DC/DC uses power that is charge current, but not measured as such going into the battery.

[arber333]

Do you have a dedicated current sensor for the CHAdeMO lines? Or an overall sensor for measuring the current going into your pack?
In case of the second option: are there any loads consuming power during charging?
If for example a DC/DC uses power that is charge current, but not measured as such going into the battery.

Like you can see from photo i connect at the battery with ISA sensor after this connection. So no current should be flowing on that side.
ISA is on the CAN1 and chademo is on CAN0.
No matter what i set in current request (20A, 30A, 12A) it will allways result in termination at the same time after charg innitialization.

What i am thinking of is multiplying "Sensor.Amperes" by *-1. Would that change anything in the overall chademo report? Obviously my VCU is requesting -20A... But if i enter request for AMAX=-20 i would actually get 240A

[larsrengersen]

Sorry, did not look closely enough.
You are indeed reporting -.
With a LEM sensor this would be a sensor orientation issue, but with IVT not likely. But yes, then I'd say multiply measured current with -1

[Bigpie]

I've got my ISA on the negative side, DCDC not running. My shunt reports positive for consumption and negative for charging.
My log shows the current as -50 amps and the EVSE reporting 50 amps. It charged for 5 minutes without issue. Now I'm not sure why.

SuccessLog.txt

(425.07 KiB) Downloaded 341 times

[larsrengersen]

It all depends on how the current deviation is being calculated. If it is using an absolute value or not for example.

[arber333]

I've got my ISA on the negative side, DCDC not running. My shunt reports positive for consumption and negative for charging.
My log shows the current as -50 amps and the EVSE reporting 50 amps. It charged for 5 minutes without issue. Now I'm not sure why.
SuccessLog.txt

I tried on another EVSE and this time i wouldnt even get contactors pulled. Just error 32! See csv file and serial report.
Strange... I will now turm ISA around so it will sense current the other way. No problem doing that...

[larsrengersen]

Is the timing of the messages you are sending correct? Data update rate needs to be 100 ms for each message.
And do you know what the 0x208 and 0x209 messages are?

[arber333]

Is the timing of the messages you are sending correct? Data update rate needs to be 100 ms for each message.
And do you know what the 0x208 and 0x209 messages are?

Yes i feed data at 100ms.
Still cant pull contactors in. Just error 32.
I have no idea about 0x208 and 0x209. They originate from EVSE which behaves really different than the first one.

[Bigpie]

If you do DBG=3 your logs will have more info

[larsrengersen]

Looks like from the beginning, the EVSE says 'Stop'. The question is, why?

[arber333]

Looks like from the beginning, the EVSE says 'Stop'. The question is, why?

Yes, but another EVSE starts charging without any problem. Only later it complains about current mismatch.
Second EVSE seems to be newer combo 22kW AC, 50kW Chademo and 50kW CCS station.