Odd rollback effect - Ideas?

[Kitfox]

This is not a complaint or decree that something's broken. Just trying to get feedback and see if anybody else has encountered this or has ideas on it.

The M3 is parked in a rather narrow garage and regen is set to Hold. The vast majority of the time when the car is moving and the accelerator is released, it does a stop. It is going in the direction, it slows, and then it stops and holds.

On rare occasion, and only in the garage when backing out, it will slow, then roll forward a few inches before going to hold.

One thing I know it isn't is ground angle. The garage floor has a 0.3 degree tilt that I'm backing barely downhill on, and then a half inch drop from the garage foundation to asphalt that has more tilt downhill behind me.
When it happens, it doesn't care if the front wheels are straight or turned.

One suspicion I have that it might be is proximity. When backing out, the rear side sensors are triggered rather close by the frame of the door and the turning angle brings the nose of the car to within 13-17 inches of the side wall of the garage. However this doesn't seem to make sense, since it doesn't do the same when going forward and coming to a stop pulling into the garage, even though it gets to within 12 inches and just has STOP displayed. The video I have of the occurrence had 21 inches at the front and the back end of the car outside the garage door.

I'm interested to know if anyone has any ideas on the cause of this, or if anybody else has encountered it. If not, I'll just ask Tesla support peeps.

 

[Kitfox]

Finally figured this out I think. When the system is rather cold, it sends power to the motors to create some heat to warm things up for the batteries and motors, and "Not enough power to make the car move". However it is a tiny bit of forward torque that's normally defeated by the Hold brakes, and when moving backward on a relatively-flat surface in hold mode, it'll cause an inch or less of forward movement before the brakes kick in when removing reverse accelerator application.

Since the car effectively just departed the inch of space it's going into, there is no concern, but it can be confusing if you're not aware of it.

 

[bedoig]

I've noticed this too when backing into my garage, but never thought there was any correlation to temperature. I thought it was just the small reverse torque being applied (on top / after regen goes to zero at low speed) to bring the vehicle to a halt before applying the breaks.

 

[Kitfox]

I've noticed this too when backing into my garage, but never thought there was any correlation to temperature. I thought it was just the small reverse torque being applied (on top / after regen goes to zero at low speed) to bring the vehicle to a halt before applying the breaks.

Then I may be wrong about the theory. I noted that when the Hold is on, there's a tiny, tiny length of power consumption showing on the screen. It doesn't happen most of the time for me. If you're seeing it in situations where it's not cold and after driving a lot, then my theory is likely to be wrong and I'm back at no idea for the cause.

 

[bedoig]

Then I may be wrong about the theory. I noted that when the Hold is on, there's a tiny, tiny length of power consumption showing on the screen. It doesn't happen most of the time for me. If you're seeing it in situations where it's not cold and after driving a lot, then my theory is likely to be wrong and I'm back at no idea for the cause.

No, I think you're generally correct. There is power being consumed. I just think it's providing opposite propulsion for braking rather than for purposes of heating. From CAN data I'm pretty sure the inverters (front and rear) can actively provide 3-3.5 kW of heating without creating any motor torque.

Edit: To clarify maybe, I think this is unique to the Hold driving mode which gives one pedal driving to zero MPH. Actual regen has historically died around 3-5 MPH until the switch from induction to permanent magnet motors and more recent software. The theory is either the PM motors overcome an energy barrier at low speed where regen costs less power than it generates, or more likely, that they're actively applying some negative torque to the PM motor to come to a complete stop. Either way, you can see the power consumption on the screen (and via CAN) and the one-pedal-to-zero-MPH mode is only available on cars with a PM motor (Y/3 and raven S/X).

 

[Kitfox]

No, I think you're generally correct. There is power being consumed. I just think it's providing opposite propulsion for braking rather than for purposes of heating. From CAN data I'm pretty sure the inverters (front and rear) can actively provide 3-3.5 kW of heating without creating any motor torque.

The inconsistency is what makes me wonder. It doesn't happen to me most of the time. Normally I put it in reverse, it's in Hold, and there is no power shown. I back, pause, and there's no roll forward before the brakes engage. But sometimes I put it in reverse, and it's in Hold mode with about a millimeter or less of power draw showing on the screen. That lets me know it's going to happen. If it happened all the time, that would be one thing, but only sometimes is weird. I like to understand things, so it piques my curiosity.

 

[bedoig]

I had a slow edit above to add more detail on my thoughts. I could totally be wrong, but it's always been very consistent for me. It always happens at the very, very slow last couple inches backing into the garage where I'm barely feathering the throttle trying to ride that line between complete stop and slow back up which is a little harder in hold. I also live in TX, so I don't really have a lot of thermal variation. Lots of weird little things to question with these cars is part of the fun

 

[Kitfox]

Near the end of backing... Hmmm... Are you in the "red zone" for the proximity sensors at the time? My garage is rather tight, and depending on how I park can have proximity sensors in the "Stop!" zone. I wonder if it might be related to something akin to 'pulling away from something that might otherwise be struck if the car keeps rolling a little'.

 

[bedoig]

Hmm, ya, true... I'm always in the red or stop zone on the parking sensors when I see this.