Leaf full stack not reversing properly

[J0hannes]

Finally had time to test this out again as I have witnessed this happening on multiple Zombie fw versions. I think it was 2.05 that I first tested, but didn't look deeper into this then.

My setup is a 2017 Leaf full stack and LBC. Leaf accelerator pedal.
Forward direction this really wants to move, and I can press against the brake pedal to verify that motor effective torque is ~260Nm reported by inverter.

Things don't work like that when drive direction is reverse. Torque request but the effective torque is max. -17Nm --> this is no good...

I don't have a manual gearbox, so I can't overcome this issue just by selecting reverse from manual box

Looking at the CAN trace, the JoysticGearPosition in 0x11A doesn't change regardless of drive direction selection. Got me wondering, could this be the root cause. Probably another safety layer by Nissan, that the inverter won't go full negative torque when drive direction selected forward.

Attached params and CAN trace.
Tested on latest released and also built todays situation of what is the most current in Damien's GitHub. The trace is done with this version I compiled.
But note, that this doesn't change the behavior of the inverter.

[tom91]

You ever had a version where it worked better?

Will have to dig into this when i get the chance to review some Leaf CAN logs. Yes the leaf does indeed have the reverse signal, can review how to add it in properly into the CAN control in Zombie.

The Leaf is indeed firmware limited in reverse.

It would be good to know if others have the same, or ever had it working well in reverse.

[J0hannes]

No, hasn't worked in any version for me at least. All versions fwd direction just ok, and reverse barely moving.

Yes, it's still two different things to have top speed in reverse limited and whatever this is, what I am experiencing:(

I think max torque should be more close, if not perfectly symmetrical forward and reverse up to the max rev limit. The car should be capable of going up the same incline forward and backward.

I did ask Raine who's Jeep Patriot is Leaf+Resolve Ev and I'm paraphrasing, there's no difference on torque forward and reverse.

[Bratitude]

yup in my datsun 720, reveres has very little torque. barley will move the vehicle

[tom91]

https://github.com/damienmaguire/Stm32- ... c08cf829e2

I had a few minutes to throw this together for testing. Please try this and let me know how it goes.

Now updated the message to reflect the selected direction, before when not in charge we always say we are in drive.

[J0hannes]

https://github.com/damienmaguire/Stm32- ... c08cf829e2

I had a few minutes to throw this together for testing. Please try this and let me know how it goes.

Now updated the message to reflect the selected direction, before when not in charge we always say we are in drive.

Got the test branch compiled and tested. This time no change in behavior.
Also, params attached

[tom91]

Can you test with applying less torque request in reverse. To see if it goes above the -17.5Nm?

For example we can see a -35Nm request form a leaf 60kWh log.

[J0hannes]

Here's one frame with -20.25Nm torque req and MG eff torque does not go above -17.5Nm

[tom91]

Here's one frame with -20.25Nm torque req and MG eff torque does not go above -17.5Nm

Please provide a log of this, I need data to review.

[J0hannes]

Here you are, I meant to attach this already in my prevvious post

Rev_trace_040325.zip

(86.32 KiB) Downloaded 491 times

[J0hannes]

These are active DTCs that Leafspy reads.
Haven't checked all of these, but some are CRC errors

[tom91]

Here you are, I meant to attach this already in my prevvious post

Your logs are not savvy CAN compatible so I cannot read them to analyse.

Screenshot_2025-03-04-21-37-52-57_dbdbb72d9b105f7d3ea161eb17f487e2.jpg

These are active DTCs that Leafspy reads.
Haven't checked all of these, but some are CRC errors

Interesting, could you please provide details on each?

[J0hannes]

Rev_040325_2210.csv

(461.68 KiB) Downloaded 449 times

Does this work?

I´ll look up the DTCs, what they are

[Marat]

P318E CAN ERROR
(CAN data error)
Traction motor inverter is unable to receive the EV system
CAN signal from VCM

P3193 CAN ERROR
(CAN data error)
Traction motor inverter is unable to receive the
EV system CAN signal from Li-ion battery controller
continuously for 2 seconds or more

P3244
DRIVE MOTOR A INVERTER
(Drive Motor “A” Inverter Voltage
Sensor Performance)
High voltage detection value of Li-ion battery
controller differs largely from that of traction
motor inverter for 10 consecutive seconds or
more

P324A DRIVE MOTOR A INVERTER VOLTAGE
(Drive Motor “A” Inverter Charge Error)
Precharging does not start for 10
consecutive seconds or more
when starting EV system

The errors may be old and need to be cleared and recalculated.

id11A id1d4 revers bit ?

[Marat]

id1F2 bit0 chrge ?

id1D4 bit0 x88 bit1 x88 (GEN2 GEN3) ... bit5 x44 dr x42 rev
dit0,bit1 Torque control, torque limitation - not yet studied.
Taken from the log GEN4

[J0hannes]

PDM

PDM

PDM

Battery

Battery

Battery

Inverter

Inverter

Inverter

Inverter

I didn't clear these yet, so can't say if they are all relevant

[J0hannes]

Well, I think I found out, why the inverter gives only -17.5Nm out...

Because Zombieverter tells the inverter to give only that

What I did was update my DBC from Dala's Github, my previous one was from over a year ago. New signals here No wonder finding out the issue felt difficult and now the problem is clear as the sky on a freezing day

Also, I guess the Inhibitor pos and brake pedal pressed statuses would be a good idea to change so they follow actual brake pedal and drive direction selected states

I think with this 1D4 getting fixed, Leaf reversing will work like it should

[tom91]

Okay thanks for finding this, I will also get new dbcs mine are old too.

I will update this post once I have update the code on github.

//EDIT///

Okay so made both limits the same lets see what that does. Make sure not to break anything.

https://github.com/damienmaguire/Stm32- ... 6c70b2aa69

Possibly this will also allow us to run the motor in reverse, but not sure just yet.

!!! CHECK your params after this update, as we are removing a request limit. This means if the VCU asks for -600Nm in reverse it the inverter might comply.

Check:
1.Regen settings
2.Ramp settings
3.Reverse throttle limits
4.Check brake settings

[Rustybkts]

It would be good to know if others have the same, or ever had it working well in reverse.

I also am just testing my full Leaf stack and reverse barely drives over a matchstick.
Has this been rectified?
I will do more testing today but cannot see any parameters I have set that would limit power in reverse only as forward is fine.

[tom91]

Has this been rectified?

I made a change as per my pervious post that needs testing. It is explained in the post.

[Rustybkts]

As a bit of a newbie regarding compiling the github files, is there a description of the steps required on the forum as I have extensive experience with numerous PLC and other industrial software packages but very rusty using the coding on Zombie.

I could help moving issues along with the Leaf full stack as it does make for a cheap and attractive all in one drive unit.

[tom91]

I can help with testing and providing the required files to complete the testing.

How are you with CAN logging? As there is no point spending time when I cannot get the required feedback to move this topic forward.

[Rustybkts]

No problem with CAN logging and want to try the update earlier posted but haven't compiled for the Zombie, only updated with the latest firmware V2.20A.
Are there any threads to follow to assist?

[tom91]

I will reach out to you in directly. Testing is only for those who can provide feedback for development or ones who can compile and code.

[J0hannes]

Yeah, sorry about the radio silence. It'll be until Sunday before I have the chance to get back to testing this.

Thanks Tom for providing a fix to test. I'll also report once I get on to testing this.