Leaf motor not spinning after update 4.77r -> 5.24r

[DarkwingDuck]

Hello all, it has been quite a long time since my last post. I finished the hardware part of my build a few months ago (leaf motor with Johannes gen2/rev3 board and i3 battery in mazda rx-8)

I got the car running and driving with the "factory supplied" 4.77r firmware and basic parameters. I decided it was time to update the firmware, which at the time was 5.24r. This was done successfully, together with Zapateros parameters, modified for my car (I use a different throttle setup).

The inverter now starts and communicates as usual but won't spin the motor. I just hear a PWM whine which increases with throttle application. Throttle works exactly as before according to logs. Resolver angle is more or less static with only small fluctuations. il1/2 readings increase with throttle application.

I tried reinstalling a few older 4.XX and 5.XX firmware versions, but the problem remained. I am currently running the new 5.27r. I have also tried Johannes own touran parameters which are more or less identical to Zapateros anyway. I have also tried lowering il1/il2/udc gains to my original values.

I have attached my current parameters and an old screenshot of the parameters from 2020 from my first test spins with the motor with a lower voltage pack. Notably "curkp" and a few other parameters are now missing and can not be added manually via command line, inverter responds that the command does not exist.

I am thankful of any and all help

[Ev8]

Does resolver angle change when turned by hand?

[DarkwingDuck]

Yes

I raised the rear of the car and spun the wheels by hand in first gear. See image "manual turn". Feedback is rather choppy as expected but angle changes when the wheel is turned.

I tried giving some throttle in neutral. Just the slightest nudge in angle with every throttle input, then the output shaft bounces back again to the original angle. See image "acc press"

[Ev8]

Long shot, are you using can for throttle? Johannes added an additional can bus speed to the options recently meaning that the first option is no longer 500kbs so need the correct speed selecting after
updating.

Ps do you have a built page anywhere?

[DarkwingDuck]

The throttle is the usual 5v deal (original rx-8), not CAN. Throttle works exactly as before according to logs.

Build page has been down prioritised in favor of building I have two years worth of build pictures and data ready though, I hope I will get around posting it soon...

[arber333]

The throttle is the usual 5v deal (original rx-8), not CAN. Throttle works exactly as before according to logs.

Build page has been down prioritised in favor of building I have two years worth of build pictures and data ready though, I hope I will get around posting it soon...

I had to replace throttle pedal on my friends rx8 with more common prius pedal. That pedal was still connected to ecu. We only tapped into signal line.
In the middle of the drive the car would decide to open throttle suddenly. I traced this to rx8 ecu automatic throttle control in neutral.

[DarkwingDuck]

Ok. I have the throttle wired directly into the inverter, totally cut off from the vechicle ECU. The Vehicle ECU still remains intact though. I will feed it with basic parameters later to get power steering etc fully functioning, as others have done. I have an Arduino DUE with CANDUE addon, managing and moderating the communication between the "old" and "new" car.

One more note. I can still get the inverter to trip overcurrent by flooring the accelerator/throttle for a few seconds. Maybe that helps in troubleshooting.

I must point out that the vehicle was fully operational and driveable before the firmware update. What could have gone wrong? Is there any other parameter I could/should log and provide to the forum? Was the 4.77 to 5.24 update a "too big leap" which has scrambled some internal parameters? I am out of ideas.

[Ev8]

Check out my build page I’m pretty sure I posted the code I us on an arduino to replace the ecu and still have functioning dash ect

[johu]

One more note. I can still get the inverter to trip overcurrent by flooring the accelerator/throttle for a few seconds. Maybe that helps in troubleshooting.

So you get inverter wine and current flow then? It doesn't really make sense but what if you add Resolver pinswap? Somehow sounds like the classic "field turns clockwise resolver measured anti-clockwise"

Software should be backward compatible down to the first version. But always new parameters are added and you need to check for sanity

[DarkwingDuck]

Pinswap solved it, see attached test burnout picture...
Thank you Johannes!

[johu]

Ok cool But also a bit odd. Are you sure all you did was upgrading the firmware? But then 4.77 is quite old, maybe sin/cos was swapped in SW along the way.