DC motor, or any other motor controlling torque or speed?

[nForce]

Are we controlling speed or torque of a motor in electric car. Or in petrol/diesel motor?

Because if we control torque, how then we control speed or opposite? Because when we have a maximum torque we have a minimum speed.

Thanks.

[T3sl4co1l]

Yes.  (The inclusive-yes, both.)

Torque or current is controlled primarily, so as to prevent destruction of the inverter, or to lesser extent, the motor or battery.

The current setpoint is then controlled by the speed setting (cruise control), with hooks to enable/change/disable the setting, and anything else that feeds into it (collision avoidance, following, etc.).

There's nothing that prevents you from hooking up one error amp to another!  Indeed, there are very real advantages to a multistage control!

Probably, in open loop (user controlled "throttle") mode, it's not just current or torque, because that would feel unnatural (an ICE's torque response varies with RPM as well as throttle position), but rather a mixture of current/torque and RPM/speed is used.  This makes it easier to cruise at a fixed speed.

Further, for induction and PMAC motors, it's not just current or voltage that's controlled, but frequency and phase sequence as well, which all need to be coordinated together.  This involves a more complicated program of feedback paths that I don't need to go into here, but suffice it to say that, on a superficial level, it's managed in a similar way.

Tim

[Benta]

To clarify:
You cannot "control torque" in a meaningful way.
Torque is something that is demanded by your mechanical load, and your drive system should be able to deliver this torque within its design limits.
Torque can be limited, of course.

For a control strategy (normally position, speed, acceleration), torque is pretty useless.

[Dave]

To clarify:
You cannot "control torque" in a meaningful way.
Torque is something that is demanded by your mechanical load, and your drive system should be able to deliver this torque within its design limits.
Torque can be limited, of course.

For a control strategy (normally position, speed, acceleration), torque is pretty useless.

Spreading misinformation yet again, are we, Benta?

Speed control of a vehicle is done by reading the speed from the output shaft and demanding a torque from the torque regulator to attempt to reduce the speed error to zero.
Speed control loop -> Torque control loop -> Current control loop (with a couple of other things)

[rrinker]

 Tim's answer hints at why CVT's feel unnatural, at least before they started programming set 'gears' into them. The feedback did not match the user input. I remember test driving a car with one a number of years ago - to someone used to always driving manual transmission, it felt like the clutch was worn and slipping, far more so than a regular automatic. Just - weird.

[max_torque]

All passenger cars, when driven by a human driver, are driving in a "torque demand" mode.  The accelerator pedal position is monitored, and various "pedal maps" are applied to create a "driver demand flywheel torque". The Engine ecu is then responsible for setting the engine control actuators (throttle, injectors, ignition, turbo etc) to create that torque. Modern "torque based" engine management systems use torque as their primary control parameter, so as to be able to arbitrate between various torque sources, for example, gearbox request (to get a smooth change) or Stability control intervention (torque down or torque up etc)

The "Human" driver closes the loop, by applying a suitable pedal input to accelerate the car at a rate that suits them, to a speed that suits them.  To enable it to all feel nice, we (powertrain calibration engineers like myself) spend a huge amount of time mapping a vast number of calibratable parameters, over 50,000  in a modern engine ecu, to deliver a powertrain response that feels natural for the vehicle application (which depends on what the car is, and how it is likely to be driven, ie a big difference between say a sports car and a van for example)

If the vehicle features cruse control, then a closed loop speed control loop is enabled to function in place of the driver demand (the driver no longer has to have their foot on the accelerator pedal), but that demand is cancelled by either the cruse control being switched off by the driver, or by the driver pressing the brake pedal, or by stability control intervention.

In an ICE powertrain, driver demand is bi-directional, but the maximum negative torque is limited by the total frictional torque of the engine, as the engine cannot "absorb" any more torque than the torque required to drive it around when it is not firing.  For an EV, which has a truely bi-directional powertrain and can fundamentally absorb as much torque as it can apply, then the maximum negative torque is limited by factors such as tyre adhesion and passenger comfort. Many current EVs feature "one pedal" driving where the vehicle can be controlled with just the accelerator alone unless an emergency stop (using the conventional friction brakes) is required.

To make maneuvering easy, some EVs actually blend out of torque control into a very low gain speed control at ultra low speed (sub 1mph) where the accelerator demand actually becomes a speed demand input to a closed loop speed controller, the output of which is blended to the "torque demand" output to give a very smooth entry and exit into low speed driving (ie in heavy traffic etc)  Because of this EVs can be driven a very low by stable speed without excessive jerk or shunt.

[Whales]

> Are we controlling speed or torque of a motor in electric car. Or in petrol/diesel motor?

We control neither directly.  In an ICE: our pedal approximately tells the engine to do "more" power, not more speed or more torque.

The exact torque and speed we get as a result is then determined by many things including the engine's characteristics, current velocity and the current load (vehicle weight, hill slope, transmission, etc).

In most circumstances: pressing the pedal down further will lead to both increased speed and torque.  In certain circumstances (eg over-revving) it can actually lead to one of these things being reduced.

[langwadt]

> Are we controlling speed or torque of a motor in electric car. Or in petrol/diesel motor?

We control neither directly.  In an ICE: our pedal approximately tells the engine to do "more" power, not more speed or more torque.

The exact torque and speed we get as a result is then determined by many things including the engine's characteristics, current velocity and the current load (vehicle weight, hill slope, transmission, etc).

In most circumstances: pressing the pedal down further will lead to both increased speed and torque.  In certain circumstances (eg over-revving) it can actually lead to one of these things being reduced.

I'd say in an ICE the pedal approximately tells the engine how much torque to produce 

[Circlotron]

Torque vs rpm is analogous to current mode vs voltage mode SMPS, or a constant current vs constant voltage PSU. You need to close the feedback loop to control voltage by controlling current, or or control rpm by controlling torque.

[amyk]

If you ever get a chance to, try driving a trolleybus and then a regular diesel bus, preferably within a short time of each other. They both have a very different acceleration characteristic compared to a normal petrol vehicle.

[T3sl4co1l]

To clarify:
You cannot "control torque" in a meaningful way.
Torque is something that is demanded by your mechanical load, and your drive system should be able to deliver this torque within its design limits.
Torque can be limited, of course.

For a control strategy (normally position, speed, acceleration), torque is pretty useless.

So I cannot control current in a meaningful way?  Current is something that is demanded by my electrical load, and my power supply should be able to deliver that current within its design limits?

Or symmetrically:
So I cannot control voltage in a meaningful way?  Voltage is something that is produced by my electrical load, and my power supply should be able to deliver that voltage within its design limits?

(The V/I analogy is directly applicable of course, as a PMDC motor for example has torque proportional to current (less windage losses), and RPM proportional to voltage (less DCR).)

I'm guessing it's just a miscommunication, in the same way that we often speak about voltage source power supplies (the load sets the current) or current sources (the load sets the voltage).  It's simply a matter of which control we are using, and the control used depends on the application.

Tim

[KC0PPH]

torque is pretty useless.

In the industrial world its pretty common to have torque control on servos... You really dont get any meaningful information from it as most torque values are given as a percent of full torque and not very accurate.

[soldar]

This is a bit of a philosophical question. It is like asking "What is the true nature of a triangle?"

It is like newbies who ask "how can I put 10 volts and 10 amps through a 10 ohm resistor?"

Everything is related to everything else.

Speed is the result of forces acting on an object. You adjust speed by controlling force/torque.

Like current is adjusted by controlling voltage. You can have a current source but ultimately voltage is what makes electrons move.

Ultimately force is what makes objects move. Distance, speed, power, etc are all a result of force applied. No force no nothing.

[GeorgeOfTheJungle]

It seems to me that you're all saying ~ the same thing but with different words: power is torque times RPM. Press the accelerator and the amount of mixture that enters and burns into the cylinder increases, which in turn generates more power, which in turn pushes the piston with more force, which is turn means more torque, which in turn tends to accelerate the engine to higher RPMs. In an EV, the mixture is the current.

[soldar]

Press the accelerator and the amount of mixture that enters and burns into the cylinder increases, which in turn generates more power, which in turn pushes the piston with more force, which is turn means more torque, which in turn tends to accelerate the engine to higher RPMS. In an EV, the mixture is the current.

No. Press the accelerator and the amount of mixture that enters and burns in the cylinder increases, which in turn produces more force/pressure on the pistons.  If the piston moves there is work generated. If the piston is stalled there is no work generated.  A small explosion inside a pressure vessel produces no useful mechanical energy.  Work is the product of two things: force and distance. Power is work divided by time.

An explosion produces pressure/force. An electrical current moving in a magnetic field produces force. Not work and not power.

(Edited to correct work and power.)

[max_torque]

> Are we controlling speed or torque of a motor in electric car. Or in petrol/diesel motor?

We control neither directly.  In an ICE: our pedal approximately tells the engine to do "more" power, not more speed or more torque.

The exact torque and speed we get as a result is then determined by many things including the engine's characteristics, current velocity and the current load (vehicle weight, hill slope, transmission, etc).

In most circumstances: pressing the pedal down further will lead to both increased speed and torque.  In certain circumstances (eg over-revving) it can actually lead to one of these things being reduced.

No, well not for the last 25 years. All modern (produced since 1995 when bosch introduced them) Engine management systems run in "torque arbitration" mode. The system is finely calibrated because it is critical to get accurate engine torque control for functionality like stability control or gear shift control etc.  By "finely" i mean a total allowable error of around +1Nm at low torques (<10Nm) and around +-10Nm at full torque (which these days is likely to be 300Nm or more on pretty much any turbo charged engine.

The accelerator directly controls that torque, which is why modern engines are so smooth and nice to drive!


(On an EV of course, the eMachine torque canbe controlled much more finely indeed, with errors expected to be less than 0.25Nm at all times, and torque control bandwidths of over 1Khz, allowing advanced chassis control systems to operate, for example, real time tyre slip control etc.)

[soldar]

The accelerator directly controls that torque,

With a name like max_torque I totally have to believe what he says!

[GeorgeOfTheJungle]

Solder your logic is flawed, again. Where in Spain do you come from?

If the piston moves there is power generated. If the piston is stalled there is no power generated.

If it burns there's power.

Power is the product of two things: force and distance.

That's not power that's work, work is energy, and power is energy/time.

[soldar]

Ha! simulpost. I corrected my post and then saw yours. At least we agree on that.

[langwadt]

Solder your logic is flawed, again. Where in Spain do you come from?

If the piston moves there is power generated. If the piston is stalled there is no power generated.

If it burns there's power.

Power is the product of two things: force and distance.

That's not power that's work, work is energy, and power is energy/time.

in the context of a car engine force and distance is useful power, the rest is just heat

[nForce]

Speed control of a vehicle is done by reading the speed from the output shaft and demanding a torque from the torque regulator to attempt to reduce the speed error to zero.
Speed control loop -> Torque control loop -> Current control loop (with a couple of other things)

I think Dave and max_torque had answered correctly. But I still don't understand. If we control electric current then we control torque, if we control torque then we control speed. But why do we have then torque control loop and speed control loop if we said we control current, and with that, we control all of the other physical quantities?

[GeorgeOfTheJungle]

in the context of a car engine force and distance is useful power, the rest is just heat

But the same goes for an EV motor that's seized... more current = more heat. You wouldn't say then there's no power input, would you?

[langwadt]

in the context of a car engine force and distance is useful power, the rest is just heat

But the same goes for an EV motor that's seized... more current = more heat. You wouldn't say then there's no power input, would you?

it isn't making any mechanical power

[GeorgeOfTheJungle]

in the context of a car engine force and distance is useful power, the rest is just heat

But the same goes for an EV motor that's seized... more current = more heat. You wouldn't say then there's no power input, would you?

it isn't making any mechanical power

You've made the same error as soldar BTW... you said "force and distance is useful power" but as time^-1 is missing there it's not power.

[David Hess]

I'd say in an ICE the pedal approximately tells the engine how much torque to produce

On a gasoline engine, the throttle position restricts the air intake.  Under normal operating conditions, there is a high manifold vacuum so air volume is proportional to throttle position.  For a given throttle position, as the engine RPM increases, there is less fuel-air volume available for each piston so torque decreases as RPM increases.

You cannot "control torque" in a meaningful way.
Torque is something that is demanded by your mechanical load, and your drive system should be able to deliver this torque within its design limits.
Torque can be limited, of course.

For a control strategy (normally position, speed, acceleration), torque is pretty useless.

Sure you can control torque (usually through the current) and some systems only have torque control where only the drive current is adjusted.  The torque is proportional to acceleration and can be integrated to get velocity which is common to stabilize the control loop.  The rate of change of torque (or current) may be controlled to limit jerk or change in acceleration per unit time.