Do DC motors "mind" being PWM'd....

.... or do they "prefer" it?

By which I mean, from the motor's point of view, what's better for the speed control of a motor:

  • Being fed full voltage for very short times ala PWM, or
  • Being fed lower voltage from say some kind of rheostat arrangement?


That's a good question but I think PWM wouldn't be the dominant method of controlling DC motors if there was a significant downside. I'm not certain (maybe someone can comment knowledgeably on this) but my understanding is that the inductance of the motor coils smooths the PWM square wave so what the motor sees is pretty much a DC current with some amount of ripple.

Also, remember a rheostat dissipates the power as heat, wasting it. With good PWM you are only supplying the power the motor actually needs.

A variable voltage approach (lower volts for lower speed) would result significant loss of torque.

Well all the motors I have asked have not expressed a preference.

I once was doing a job that needed maximum reliability so I asked a fan manufacturer what was the best method for reliability and they said the variable DC voltage.

However, that was just there fan, I don't know if it applies universally.

A DC motor shouldn't "mind" a PWM supply. When you think about it, the rotor windings are switched in and out of circuit on a continuous basis by the commutator as the motor operates. So switching the supply on and off via PWM is very much the same operation.

As others have said, using a rheostat to control speed is both wasteful of power and poor speed control. Assuming a preset rheostat value, as the motor load increases, the current increases, so the voltage loss across the rheostat increases. This leaves less voltage to the motor terminals so the motor tends to slow down. (Motor speed is proportional to voltage)

On the basis that PWM voltage is averaged out, then motor speed is effectively proportional to PWM duty, irrespective of shaft load

Kind of anecdotal and maybe OT but... I was experimenting with some big (10 amp) brushed motors and I noticed that they tended to "sing" when using PWM.

they tended to "sing" when using PWM.

Yes they will, this is common. It is due to the windings being attracted / repelled as the current is pulsed through them. Some people try and push the PWM into the ultrasonic region to get round this.

It's a brief PDF, it could have been a one-pager if he'd scaled his pics for that, nonetheless a good explanation therein.

With PWM you are using the inductance of the motor windings to smooth out the current - for best efficiency there is a balance between switching too slowly (leading to unsmoothed current variations which tend to waste more heat in the copper) and switching too fast (possibly leading to “iron” losses from the magnetic circuit, but much more likely causing more switching losses in the H-bridge).

Mechanical vibration due to the windings vibrating should be reduced by higher frequencies (less variation in current as well as lower amplitude of movement possible in shorter period). For long-term reliability I suspect a low PWM frequency might cause mechanical weakening of the windings’ support (as well as annoying noise levels - )

PWM leads to electrical noise - this needs to be controlled to prevent EMI - although the alternative of varying the DC voltage would probably need a switch mode supply anyhow (which has same issues).

PWM is much kinder to the motor mechanically than just connecting the full supply voltage suddenly, note!