A synchronous AC motor controlled by a variable frequency drive unit (VFD) is a pretty common thing in industrial settings. It’s also the system chosen by Tesla to drive their Model S.
I’m wondering about how these systems are controlled. Just so we’re all on the same page, let’s assume we’re talking about the Model S’s drivetrain, in which the motor has a permanent-magnet rotor, and the VFD manipulates the current in the stator (field) windings.
I am aware that RPM in these motors is in lock-step with the frequency applied to the field windings by the VFD, and that torque is exerted when the rotor’s magnetic field lags behind the stator’s magnetic field by some non-zero angle, and that when this lag angle gets too large the motor can lose sync. But it would seem that during acceleration, the VFD wants to ramp the motor RPM up at some predetermined rate. Does this mean that the VFD also has to adjust the peak value of the current to the field windings in order to avoid the lag angle getting so large that sync might be lost?
If the Model S’s VFD controls the RPM ramp rate, then does that mean the car accelerates the same whether I’ve got one person or four people in the car? What happens if I attach a 4000-pound trailer? Surely the VFD will be forced to choose a slower RPM ramp rate to avoid losing synch with the motor, but how does it “know” to do so?
The right foot pedal is not exactly “RPM”. That would be insane…
That would be like the classic Ferrari crash problem… if in first gear, and driver pushes the pedal too hard, then it jumps forward radically , the driver is unable to work the pedals… The inconsistent and rough city street surface causes torque steer and the driver fights very hard to control steering… but this prevents fine control of the feet… which can mean that the acceleration increases… it all ends in tears.
The only way to do it… Tesla accelerator pedal sets power, and hence force, just like with internal combustion engine. (gasoline/petrol, distilate/diesel , LPG or other gas.)
If you put it down a little bit more, the computer knows that you are requesting a bit more power. The computer treats this as a request, In that the computer judges what it can give… if it can’t provide it, then it gives what it can and lets the driver wait for the extra power to come about.
What that means in practice is that if you are you doing 20, and request the power level that drives the car at 100, well it won’t run the engine at the RPM’s for 100 straight away…
If you have 4 people in a light car, then you will accelerate slower with the same power delivered at first, but then the driver will soon get used to asking for more power to achieve a reasonable acceleration .
I’m not particularly concerned with what the accelerator pedal is mapped to. I’m more interested in the interaction between the VFD and the motor itself.
Let us assume for the sake of discussion that I have mashed the accelerator pedal to the floor, and now the VFD and drive motor must work together to accelerate the car as quickly as possible. How does the VFD know how rapidly it can increase the RPM of its magnetic field without losing its magnetic “grip” on the rotor? How is it controlling torque while simultaneously controlling RPM? And so on.
I haven’t worked with these types of motors, but I would assume that the controller uses Back EMF and possibly an optical encoder to control the speed ramp.
I would imagine that the frequency is not what is the primary input to the system, but rather it is a consequence. If you have a phase angle detector - a hall effect sensor works nicely as you already have magnets, you modulate the delta phase angle that you are delivering power over to vary the input power. As the motor speeds up nothing else changes - the system still watches for the effective 0 phase (or the equivalent for multiple pole motors) of the motor and delivers power relative to that. This means that a stalled motor is provided with effectively DC until it begins to move, and the frequency ramps up in synchrony with the motor. This is why such motors have such insane low down torque.
Synchronous motors are not common in industrial settings. VFDs eliminated the need for them. In industrial settings VFDs are normally connected to induction motors.
A synchronous motor does not have any slip it gets very exciting when there is any slip. I have seen a 6000 Hp main motor bounce because it got out of sync.
I believe the motor you are questioning is really a DC pulse motor. I am not really sure but my swag is pulsating DC.
Now to your question. The motor and VFD are programed to max acceleration limit, you stomp on the pedal and if every thing is in range then you get the max acceleration. With a heavy load the amps will be higher for a given speed, a feed back loop keeps the voltage from increasing to fast keeping the amps in limits to voltage.
Sorry I know this is not the best explanation but I am tired.
From what I’ve read, the Tesla uses a 3 phase AC induction motor, not a DC motor using PWM. They use a VFD to drive it, but I don’t know the details of how it works.
For industrial use, you are usually controlling the motor’s speed. VFDs can be run with or without feedback. The type without feedback is cheaper, and either the machine has to be designed so that it won’t slip excessively over the entire operating range of the machine, or the slip has to be acceptable if the torque load gets too great. Feedback control is more expensive, but allows you to use more complex control algorithms. You can measure the speed of the motor and can make adjustments if it can’t keep up with the current mechanical load. You can also measure the phase angle of the slip to determine how much torque is being applied.
As Isilder said upthread, controlling the speed for this type of application doesn’t make much sense. This is just a guess, but I’m with Francis Vaughan on this one. I think they measure the slip angle since that gives the torque, and they try to control based on that. Push down on the “gas” pedal, and you want more torque so you’ll go faster. Release the pedal and you don’t want to stop, you just don’t want to add more torque. You want to just maintain the current speed. So instead of dropping the frequency, it just maintains the current frequency which keeps the slip angle at close to zero. Or maybe it slowly drops the frequency to simulate the way a gasoline engine slows down when you let off of the gas.
If you are accelerating and the slip angle gets too far out of whack, you know the motor can’t keep up with the requested torque and you back it off to what the motor can handle. Press on the brake pedal, then it drops the frequency and the regenerative braking kicks in.
VFD’s don’t even have to “Measure” the rotor slip angle anymore. That was more a requirement of “Brushless DC motors”. Modern VFD’s can achieve almost perfect speed OR torque control just from a tuning run to calculate the motor model.
In the Tesla, I would have to think that (since they are emulating an ICE), the motor would be controlled in a torque mode rather than speed.
And no, this doesn’t mean that the acceleration is the same for all load conditions, any more than it does for an ICE. The motor has a maximum power it can put out, and it the VFD will limit the motor from exceeding it’s ratings.
I had read some time ago in a Car and Driver review that the Model S uses an AC permanent-magnet sychronous motor.
However, I now see that this conflicts with the information reported on Tesla’s own website, which says the car uses a three-phase, four-pole AC induction motor with a copper rotor. In addition to being more likely correct because it comes straight from Tesla, this also makes more sense because a permanent-magnet motor of thsi size (as reported by Car & Driver) would have an intense (potentially problematic) magnetic field in its vicinity even when powered down.