AC charging with Foccci and Zombie, not starting

[manny]

On my Saxo project I have Zombie and Foccci for charging. I'd like to add the MG ZS 6,6kW charger to the code.

To start this I want to check the rest of the system. So I selected the outlander charger for now. Planning to use the outlander code as base for the MG ZS charger.

I checked out the zombie code and as far as I can tell it should enable the EVSE and send AC to the charger and enable the HV. Without the charger communicating with Zombie?

Used the ConfigFocci function on Zombie to set up the Foccci.

Foccci is on firmware : 0.42.76.B as it came from the OI shop.

Zombie is on firmware : 2.22.A

In the spot values the EVSE current limit is correct and updating when I change the setting on the EVSE.

Also have the CPspoof set on the Zombie and it works fine, duty cycle changes with the EVSE setting.

Found some things that are off but don't know if this is a problem:

• Cablelim and PilotLim are the pilot limit(Cablelim is not used currently I think)
• MaxTemp is mapped in the Foccci but the Zombie code labels this as ChargePort_Plug
• the AcOcbState and AllowUnlock CAN maps overlap

I have spent many hours coming through the code and checking things. But I am a bit lost. Could use some help.

[tom91]

What is the real issue now? Does Zombie not go into charge mode?

I do not really see an issue highlighted just a few small niggles of CAN mapping.

Cablelim and PilotLim are the pilot limit(Cablelim is not used currently I think)

Check the Charge Power this gets modified based on this

MaxTemp is mapped in the Foccci but the Zombie code labels this as ChargePort_Plug

Wierd should not be this wen I checked it

the AcOcbState and AllowUnlock CAN maps overlap

Probally a mapping mistake, luckily OBCstate is not actually sent that long.

[manny]

What is the real issue now? Does Zombie not go into charge mode?

Yes, Zombie does not go into charge mode

[tom91]

Your zombie go into drive mode?

[tom91]

You on this release? https://github.com/damienmaguire/Stm32- ... /tag/2.20a

[manny]

Your zombie go into drive mode?

Yes, no problem. with CAN control of the OI inverter board.

You on this release? https://github.com/damienmaguire/Stm32- ... /tag/2.20a

2.22.A will go to 2.20.A release and report back

[tom91]

Key is that ChargePort_Status has not been modified by anyone in Foccci code, unfortualitley some code changes broke things that I had working by others.

So possibly there is now a miss match between what worked with zombie.

Zombie will always try starting AC charging with Foccci.

[manny]

Check the Charge Power this gets modified based on this

This works fine. EVSE set to 13A.
Foccci reports 12A and Zombie Pwrspnt = 2592W (12*240*0.9)
and EVSE set to 6A
Foccci reports 6A and Zombie Pwrspnt = 1296W (6*240*0.9)

[tom91]

The FOCCCI Revision BigPie is running is https://github.com/uhi22/ccs32clara/tre ... b4ec408c07

[johu]

unfortualitley some code changes broke things that I had working by others

Can you pinpoint what exactly broke it?
I mainly did cleanup work and added standalone charge mode. But also did some bugfix where starting AC charging again needed a power cycle. And I may have changed AcObcState behaviour do achieve that

[tom91]

Can you pinpoint what exactly broke it?

No I cannot, I have not touched the Foccci and Zombie related code since it worked. Will have to dig into it again at some point, I will add it to the list of things to look at regarding the Zombie.

Making changes without checking dependencies does really break the the OI "Eco System", guess I would need to start making branches of verified working versions of each code base then.

[tom91]

In the spot values

Can you show all the Foccci spot values? While Plugged in and Zombie shows the EVSE limits.

[manny]

This is with the original code.

Tried to compile the firmware revision BigPie is using, but having a hard time compiling the code.

Would be nice if the ccs_clara firmware had some releases on github like Zombie or OI.

[uhi22]

The compiled binaries for each commit are available on the "actions" page. https://github.com/uhi22/ccs32clara/actions
I think the build #80 is the one which you are looking for.

[uhi22]

Ohh, wait. The screenshot of your spotvalues shows me two hints: I do not see AcObcState. And I see version 0.42.76.B. The "B" means, it is a build from from the github actions. And the 76 means, it is github action build number 76. So the conclusion is: This is a quite old software. Update to #80 or even the latest is recommended.

Edit: Ohh, wait again. AcObcState is not on the spot value page, because it is on the parameter page. I'm not sure whether this makes sense, should/could we move this to the spot values? Or keep it as test-parameter, to allow manual testing?

Edit2: Added the hints with the version number and the github action explanation to the FAQ on the Foccci wiki page.

[manny]

Did not know about github action.

Will update to #80 first an report back.

[tom91]

Foccci is on firmware : 0.42.76.B as it came from the OI shop.

What is the hardware version and when did you purchase it?

I completely did not recognise the fact the 76B is the build identifier. So we might have concluded it does not longer work too early, best test a release as you are doing now to see if it works as it should. Then you can update to the latest to see if it still works.

[manny]

Updated to #80 without issue. The charge behaviour is the same.

weird bonus: had trouble connecting to Zombie via CAN but after the update it is instant

What is the hardware version and when did you purchase it?

Ordered on 17 dec 2024 it a version 5.1

[tom91]

You are not running a Lock?!? That got fixed after build 80

[manny]

Have a lock. But accidentally bough three wire type. So one wire of the lock switch is connected to one of the motor wires. have no feedback at the moment.

[tom91]

have no feedback at the moment

So Foccci does not know you have a lock right now.

[uhi22]

Would the AcObcState help to narrow the search window?

Edit: Regarding the Lock: The plan was, that for the case of no lock feedback it is possible to set both detection thresholds to the same value (zero IIRC), and then the lock is just controlled time-based, and Foccci assumes the intended lock state is reached after this time.

[tom91]

Edit: Regarding the Lock

Yes this is implemented after 80B

[manny]

I tested the lock with the auctuator test. That worked.

When i plug in the charger. The lock does not operate.

Regarding the Lock: The plan was, that for the case of no lock feedback it is possible to set both detection thresholds to the same value (zero IIRC), and then the lock is just controlled time-based, and Foccci assumes the intended lock state is reached after this time.

so this is not implemented? I can set "Lock feedback value" to test if this makes a difference

Yes this is implemented after 80B

Shall I update to latest version?

[tom91]

Shall I update to latest version?

Go for it.