Very very easy. Hold the hockey puck style stalk to the left for a two seconds, or simply put it in 'R' when you are driving. It goes to Neutral. I don't ahve the courage to put in 'R', so I hold it the left. Have done that many many times. Also going back to D is just putting it in 'D'. Same motion.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Very very easy. Hold the hockey puck style stalk to the left for a two seconds, or simply put it in 'R' when you are driving. It goes to Neutral. I don't ahve the courage to put in 'R', so I hold it the left. Have done that many many times. Also going back to D is just putting it in 'D'. Same motion.
I'd wager he meant that it's not easy to do inadvertently, as is apparently the case with the Model 3.
So I played with this when driving this morning. Two things. First, it took a while before I managed to get it into neutral at all. If your foot is on the accelerator or you press up too far to reverse it just beeps at you and does nothing. And when I did manage to get it into neutral, pressing the stalk down the same way I put it into drive in the first place put it back into drive.
Thanks for confirming that. That's exactly how every Tesla car works.
hacer
Active Member
I was just playing with this, and my car (now) behaves as described - it's a bit difficult to get into neutral while moving. The behavior has changed from the original software that shipped with my car. I'm now on 2018.4.9 700fb88. With this version, to get to neutral while driving you must push up on the stalk to the first detent (not second - that just beeps at you) and you need to hold it there for ~ 2 seconds after which it goes into neutral. Very hard to ever do this by accident, and for me more difficult and slower than I'd like for emergency use on ice for a rear-wheel drive vehicle. Simply push down (I think to the second detent) and you're back in drive instantly. It is much easier to put my model S into neutral while driving, just one bump up on the stock and you're there.
hacer
Active Member
That is always legal, but less efficient. In a model S with the induction motor(s), for neutral, it completely shuts down the drive signal and the motor(s) free-wheel. But when you keep it in drive with the throttle in the "sweet spot" between power and regen, the inverter instead creates a synchronous field with zero slip between rotor and stator. It does that so that it is smooth and continuous with no steps or bumps as the load and/or accelerator pedal changes. But to do that, the inverter needs to generate a rotating field and so it still has all of the switching losses and some of the resistance loss in the stator coils and driver. Neutral avoids those losses altogether and ensures "perfect" coasting which your foot will never get exactly.
For the model 3, with its PM motor, again neutral can simply open-circuit the stator drive and have no switching or resistive losses (other than eddy currents (which are small but omnipresent when moving). In drive, the inverter continues to generate a synchronous field that matches the back-EMF so that, once again, everything is smooth and continuous between power and regen with no steps or bumps.
While most (all?) states outlaw coasting downhill in neutral, they generally do not prohibit the use of neutral on level or up-hill roads. Tesla's coast beautifully on level (or ever-so-slightly downhill) roads. Optimal hypermiling is achieved by accelerating slightly above the desired speed then coasting until slightly below it.
Not sure how Tesla does it, but keeping the field up is not needed to provide smooth mode shifting (as demonstrated by shifting back into drive from neutral). Assuming the motor receives a torque command:
The PM motor has direct positional feedback, so the inverter knows where phasing should be for acceleration or regen and can drop field strength to zero when the request is for 0 torque.
For AC, slip and field strength determine torque. Motor speed is known (or able to be sensed), so it can drop field strength rather than keeping the field right at motor speed for 0 torque.
hacer
Active Member
Upon further testing it actually goes into neutral (in terms of free-wheel rolling) as soon a you push up to the first detent. It will go immediately back to drive if you release before the ~2 second timeout. Once you've held it there long enough, the display shows N and it will stay there until you push the stalk down to go back into drive.
So it's not a safety concern for emergency use on icy roads because you get the free-wheeling instantly, and there is no risk of accidentally getting into neutral.
hacer
Active Member
I have no doubt that Tesla modulates the field strength with the inverter, however very high power switches have limits to how narrow a pulse they can produce so there is a lower limit to field strength below which it is more efficient to have a larger field strength and use the phase (or slip) to control the torque with high resolution. I don't know what Tesla does, but at the zero-torque position with small +/- torques that seems like it would be the smoothest way to control the power without having any step-wise jumps around zero (especially at speed where the drive frequency is not low). That would mean that feathering the throttle to hold it to zero torque would result in some inverter/motor losses that are eliminated if you simply turn off the field altogether. My experience is that using neutral to coast (on nearly level road) produces lower energy consumption than I can achieve with pedal feathering, but I also admit that I'm not able to feather perfectly.
Quite possible they do it that way and keep the field up to monitor rotor speed.
General discussion:
Interesting point regarding the low end of the PWM range. I'm used to BLDC drives which care little for waveform shape, but an AC induction motor greatly prefers a sinusoidal drive signal. To create a low amplitude version low frequency version would require very small pulses at the edges.
That said, due to turn on/ turn off time, I'd expect the logic level controller to be able to create pulses so short that they never fully enable the switching device. Above that level, it is the PWM resolution/ frequency that determines pulse width. The frequency needs to be high enough to handle the top motor RPM with a decent wave shape.
Creep/ summon seem like they would be the worst case for min pulse width: very low speed, very low torque.
Similar threads
