What limits the no-load speed of a motor.

Buk___

10 kW
Joined
Jul 28, 2017
Messages
750
Take any motor and run it at 36V with no load and it will achieve some speed. eg. 300rpm (~36kph for a 26" wheel) giving an kV of 8.33

Now apply 48V and run it no-load again and it might do 400rpm (48kph).

So what stops us just continuing to up the voltage to achieve ever greater (no-load) speed?
 
Heat is always our motors' limiting factor. Power is torque X rpm. Current creates torque and the heat from current is primarily from copper losses, which is resistance X current squared. Heat is also created on the rpm side in the form of iron core losses. The alternating magnet field creates heat in the stator steel and to a lesser extent in the magnets, and it increases with frequency, so the faster those alternating pole magnets pass the stator teeth and the faster the current changes direction in the copper windings the more heat is created. While the iron core losses increase in a generally linear fashion with rpm, as opposed to copper losses that increase geometrically with current, the iron losses occur whenever the motor is spinning whether you're on the throttle or coasting down a hill, so the heat accumulates continuously.

Since the wiring and magnets in our motors are limited to relatively low temperatures compared to the parts in gas burning engines, how quickly we can dissipate the heat created in the motor to the outside world is especially important. That's why you see members going to such extremes to modify our motors.
 
You can increase the no-load speed by increasing the voltage as much as you want. Eventually you will reach a speed where something flies apart. Iron and other losses will become large above a certain point and you lose efficiency, so there's a sweet spot you want to stay in. There are also limitations in the controllers in most cases that limit how fast you can switch those FETs.
 
John in CR said:
Heat is always our motors' limiting factor. Power is torque X rpm. Current creates torque and the heat from current is primarily from copper losses, which is resistance X current squared. Heat is also created on the rpm side in the form of iron core losses. The alternating magnet field creates heat in the stator steel and to a lesser extent in the magnets, and it increases with frequency, so the faster those alternating pole magnets pass the stator teeth and the faster the current changes direction in the copper windings the more heat is created. While the iron core losses increase in a generally linear fashion with rpm, as opposed to copper losses that increase geometrically with current, the iron losses occur whenever the motor is spinning whether you're on the throttle or coasting down a hill, so the heat accumulates continuously.

Since the wiring and magnets in our motors are limited to relatively low temperatures compared to the parts in gas burning engines, how quickly we can dissipate the heat created in the motor to the outside world is especially important. That's why you see members going to such extremes to modify our motors.

I think you may have missed that I was asking about no load speed.

With no load, the motor draws very little current, so produces very little heat, but in the real world with a small motor, there is -- or at least seems to be from practical experiments by many people -- a limit above which increasing voltage fails to increase the no-load maximum according to the kV multiplier.

I tested this with motors from cooling fans and disk drives; and there are several examples posted here recently where increasing the voltage fails to increase the no-load speed by the expected amount.

My "understanding" of this is that once you increase the electric field within a core to the point were the magnetic flux exceeds the knee of the B-H curve of the core material, the linearity of the the V to rpm no longer applies.
 
John is correct, *core* losses are proportional to frequency only and irrespective of load. As he says, they are present even when coasting (no load). Try pedalling a large hubmotor even when the phase connections are open circuit - very difficult because it's turning lots of your pedalling effort into heat. Some lightweight, high speed motors will also cook themselves running no-load on a bench because they're generating several hundred watts of heat in core losses (and double-digit copper losses).

You should see a roughly linear relationship between applied voltage and motor speed because you are balancing the back-EMF, which is proportional to speed. In reality it won't be quite linear as windage will apply a non-linear load and at some point, applying, say, 1V more than the BEMF isn't going to allow sufficient current to flow to provide enough torque to overcome that load.
 
Buk___ said:
I tested this with motors from cooling fans and disk drives; and there are several examples posted here recently where increasing the voltage fails to increase the no-load speed by the expected amount.

My "understanding" of this is that once you increase the electric field within a core to the point were the magnetic flux exceeds the knee of the B-H curve of the core material, the linearity of the the V to rpm no longer applies.

You would have a hard time reaching core saturation with no load on the motor. This is a function of current, not voltage.

Motors that "fail to increase by the expected amount" probably have some other factor involved, like the controller. A motor will increase in speed until either the BEMF matches the source or the power input matches the combined losses+load. With no load, there are still losses.

Take a series wound brushed motor with no load and see what happens when you apply power. It will go faster and faster until it flies apart (this is quite violent when it happens) unless your power input is so small it can match the losses before it reaches the grenade RPM.
 
I somehow never damaged a car starter motor testing them with jump leads and a starter battery. I had no idea what I was potentially doing to them! :D
 
Back
Top