Notes on Zombie v2.2a update

[Rustybkts]

Just installed the V2.22a and confirmed its install version on spot values page.

LIN fix for VW heater works brilliantly but the reversing power fix that @tom91 gave me to try hasn't been implemented on this version and again the car cannot run over a matchstick in reverse.

Leaf stack still draws too high a current on the permanent battery supplies to the inverter and PDM. How does Damien reduce this on his cars as the current draw on these wires is too high and drains a battery in a day or two.
Connecting 12v to the same terminals on another Leaf stack gives a similar current drain of 560mv inverter and 280mv for PDM. Adding these to the Zombie and Orion "always on" and it soon depletes the battery.
I can switch these but it causes issues when charging using the PDM with the ignition off. I wonder how Nissan do it?

Two more (humble) requests for myself is for the HV request input to be fixed to allow Chademo control by Orion BMS and for 16A plug detection, neither of which is urgent.

[tom91]

So you are saying your Nissan Leaf stack still sends CAN frames after the zombie told it to sleep? What version of the leaf stack do you have and have you wired the ignition relay as required.

What is the current draw, and how long have you measured it for?

Nissan reverse fix is out of the release, things take a while to get tested. The Testing branch is 45 commits ahead of the release.

[Rustybkts]

So you are saying your Nissan Leaf stack still sends CAN frames after the zombie told it to sleep? What version of the leaf stack do you have and have you wired the ignition relay as required.

The current drain from the "always on" terminals of the Leaf stack has always been an issue to my project and I purchased a mA clamp meter to monitor it easier.
Its a Gen 2 2015 stack and I purchased another complete stack just to prove it wasn't a fault on the one installed.
I am not saying the stack is sending CAN frames but just that the always on connections never go to sleep.
The ignition relay wiring and in fact all wiring has been checked numerous times and correct. Current increases with ignition on.


What is the current draw, and how long have you measured it for?

560mv inverter and 280mv for PDM. It never changes with ignition off.

Nissan reverse fix is out of the release, things take a while to get tested. The Testing branch is 45 commits ahead of the release.

Understood.

[tom91]

Post CAN logs of your setup:
1. Off 12V permanent on
2. Zombie going into drive
3. Zombie going off
4. Wait for CAN traffic to stop

and measure the 12V current draw for each.

When you say 560mV you mean mA?

[Rustybkts]

Post CAN logs of your setup:
1. Off 12V permanent on
2. Zombie going into drive
3. Zombie going off
4. Wait for CAN traffic to stop

and measure the 12V current draw for each.

Will do

When you say 560mV you mean mA?

Oops yes

[Rustybkts]

Edit:- Tab and blank spaces not allowed so messed up my table below.

CAN logs and fixed position screen pics attached.
Current measurements only taken on permanent supply wires as ign on supply to inverter and PDM have been ignored for the test.
Inverter PDM
1. Off 12v permanent on 12v always on ign off 574mA 294mA
2. Zombie going in to drive 12v always on ign on 580mA 351mA
3. Zombie going off 12v always on ign off 2 575mA 294mA
4. Wait for CAN traffic to stop 12v always on ign off sleeping 575mA 294mA Messaging doesn't stop.

CAN traffic were already running after the car was left overnight. Does the Zombie ever stop sending CAN messages when its left permanently powered on pin 56 as required?

[tom91]

It should do. That will be your problem, will dig into it when I have time. This was working for others and myself on the bench.

Provide your params. What else is wired up right now? As those are not Zombie frames for the leaf stack.

My guess is your BMS does not stop sending CAN frames and this upsets the Leaf stack. Frame IDs look like SimpBMS

[Rustybkts]

Ok great, might be getting somewhere.
It is an Orion BMS and will be sending CAN messages as it is left running (140mA) to allow charging with ignition off as there doesn't seem to be an output from Zombie to wake up the BMS when the charge plug is sensed.

Will stop the BMS and see what happens.

Edit:- the PDM drops to zero quickly and all CAN traffic stops when the BMS goes to sleep but the Inverter permanent supply still sits at 580mA.
This isn't a major issue as the inverter permanent supply can be switched using the ignition relay. Only the PDM is needed to operate without the ignition on.

[Rustybkts]

Had a long think on this and cannot see how to enable the BMS when the Zombie detects a plug is inserted.
I cannot use the HV contactors as a trigger as it needs the BMS first.
I did try using the Water pump output as that came on instantly as the plug is inserted but the Zombie doesn't wait for the BMS and goes in to fault.
Ideally need a signal from the Zombie but that would need it to wait for the BMS CAN.

How has this been enacted with other BMS as leaving the BMS running keeps the PDM awake due to CAN messages.
A quandary others will have when using the Leaf stack.

[Rustybkts]

I managed a simple work around that may be of interest to others.

My charge port has a door and switch so I enabled the BMS when the door is open. The Zombie ECU is happy with this and the PDM enables as soon as the BMS CAN messages start.
Simple diode logic from the ignition on rail enables the BMS when in run mode.
I switched the always on power to the inverter from the ignition on rail.

Zombie ECU seems happy with this arrangement and the vampire drain is now a controllable 180mA.

Using Rev 2.31TH that I was sent for testing as it has the reversing power fix.

[tom91]

So you wired the BMS up/configured to only send CAN messages when in Run or Charging mode?

[Rustybkts]

So you wired the BMS up/configured to only send CAN messages when in Run or Charging mode?

Yes. The Orion has an "always on" supply that takes a negligible current but wakes up immediately when its power on is connected.
I hadn't realised that the Leaf PDM was staying awake due to the BMS on the same CAN. I suppose on another project this could be kept separate with the PDM on the other CAN.

It was easier just to power the BMS when the Charge port is opened and also run it via a diode from the ignition on rail.

Running both the always on and ignition on Leaf inverter supplies from the ignition on rail causes no issues as stopping CAN messages didn't reduce the always on power and I now only have the Zombie powered up with minimal battery drain.