Brushless Sinusoid Commutation

[dknguyen]

How do they produce the sinusoid to commutate a brushless motor? I understand the square wave stuff...it's pretty simple flick a transistor on and off. But how do they make the sinusoid? I'm sure they don't turn the MOSFET part way on or use resistances or something...too much energy wasted.

 

[Nigel Goodwin]

They don't, just as they don't in a brushed motor - you simply switch the coils ON and OFF - why would you want a sinewave?.

 

[Styx]

They don't, just as they don't in a brushed motor - you simply switch the coils ON and OFF - why would you want a sinewave?.

to produce a sinusoidal flux linkage in the airgap.
IF you are after minimal torque-ripple then a sinusiodal driver is the only way to do with and that is only possible if you have a sinusoidally wound field.

if that is the case then you need to synthisize a sinewave current waverform via PWM where the pulsewidth tries to synthisize a sinewave voltage

 

[Nigel Goodwin]

to produce a sinusoidal flux linkage in the airgap.
IF you are after minimal torque-ripple then a sinusiodal driver is the only way to do with and that is only possible if you have a sinusoidally wound field.

if that is the case then you need to synthisize a sinewave current waverform via PWM where the pulsewidth tries to synthisize a sinewave voltage

Have you ever seen a brushless DC motor that does that?.

 

[dknguyen]

I was thinking about what Nigel said...do you mean that they still use square waves but they have the graph of the duty cycle vs time is a sinusoid? So effectively when the motor averages out the PWM pulse it sees a sinusoid input voltage?

What I mean about using a sine-wave is the 3 coils of a brushless motor are commutated with phase-shifted "sine waves" so there is smoother, more efficient motion from one pole to the next (not sure if I'm even using the right words when I say this). Rather than a jerky motion of you just switched coils on and off in a sequence.

if that is the case then you need to synthisize a sinewave current waverform via PWM where the pulsewidth tries to synthisize a sinewave voltage

Which upon closer inspection seems to be what Styx is saying.

 

[Nigel Goodwin]

I was thinking about what Nigel said...do you mean that they still use square waves but they have the graph of the duty cycle vs time is a sinusoid? So effectively when the motor averages out the PWM pulse it sees a sinusoid input voltage?

What I mean about using a sine-wave is the 3 coils of a brushless motor are commutated with phase-shifted "sine waves" so there is smoother, more efficient motion from one pole to the next (not sure if I'm even using the right words when I say this). Rather than a jerky motion of you just switched coils on and off in a sequence.

Electric motors have got along perfectly happily simply switching the coils IN and OUT, conventional motors use brushes on a commutator to do the job, and a DC brushless motor uses transistor switches actuated by hall effect sensors to do EXACTLY the same job (except better).

I would have thought any possible advantage of trying to feed it with sinewaves would be so small as to be pointless - if you're generating variable frequency sinewaves why not just use an AC motor?.

Have you got some specific reason for this question?, because, personally, I think you're barking up the wrong tree - and I see a LOT of DC brushless motors!.

 

[hjames]

https://www.epn-online.com/page/334...brushless-dc-motor-drive-designs-summary.html

However, the rapid "square wave" switching of such a circuit produces a large amount of electrical (and acoustic) noise. This can be unacceptable in many environments; as a consequence the controller may need to include software to generate a "smoother" sinusoidal motor-coil current. The many complexities in BLDC drive design have led to component manufacturers developing integrated products to make the design process more straightforward.

Well, they do make chips which do this:

**broken link removed**
**broken link removed**

 

[dknguyen]

Have you got some specific reason for this question?, because, personally, I think you're barking up the wrong tree - and I see a LOT of DC brushless motors!.

I'm just curious so that I don't have to rule out the use of a brushless motor (due to inability to make commutation electronics) in the future.

 

[Styx]

Have you ever seen a brushless DC motor that does that?.

Oh I don't know abt 1/2 of the perminant magnet machine I have around hte lab.
There really isn't such a thing as a brushless DC machine.
A machine with a trapizoidal flux distrobution will quite happily run with a sinusoidal induced airgap flux, just like a sinusoidally wound PM-machine will happily drive with a trapizoidal flux distrobution.

Calling a machine Brushless-DC really isn't correct (like saying someone is having a heart attack). a PM-machine is either sinusoidally exited or trapisoidally exited, both of which are AC (difference being one is easier to generate)


THERE are alot of advantages to sinusoidally exciting a PM-machine, the main one is torque ripple. In a trapizoidally excited PM you can ONLY get zero torque ripple IF and only IF you can inject rated current into the winding (to produce the required flux distrobution) instantaniously, which as you know is impossible due to a small thing called inductance.

This results in commutation dips in the current which manifests themselves as torque-dips at commutation.

Sinusoidally excited PM (with optimum mag setup) only need to generate sinusoidal flux ==> sinusoidal current which is a damb sight easier then trapizoidal blocks of current.


A Sinusoidally distrobuted machine when excited with a trapizoidal control will actually give a worse torque ripple then a trap+trap

BUT a trapizoidally distrobuted machine exited with a sinusoidal control will give a vast improvement in performance, BUT you run the risk of saturating the stator iron and thus to get the best out of it it becomes a "lazy" machine

 

[Sig239]

I am currently trying to get into flying electric r/c airplanes. The motor I bought is a brushless, sensorless motor. Now you have got me interested in exactly what the output of the ESC (Electronic Speed Controller) is. Obviously a three phase set up, but I'm not sure if it is a sinewave or what. I know it contains a microcontroller and certain parameters can be programmed.

 

[Nigel Goodwin]

I'm just curious so that I don't have to rule out the use of a brushless motor (due to inability to make commutation electronics) in the future.

Don't worry about it - you just switch them, but it's most likely to all be in the motors circuitry anyway - you don't normally need to provide the switching.

There really isn't such a thing as a brushless DC machine.

Of course there is, in as much as you get brushed DC motors, a brushless one simply replaces the commutator with transistor switches triggered at the correct time by hall effect sensors. They both work in an identical fashion - with the main difference being one spins the coils, and one spins the magnet.

 

[Nigel Goodwin]

I am currently trying to get into flying electric r/c airplanes. The motor I bought is a brushless, sensorless motor. Now you have got me interested in exactly what the output of the ESC (Electronic Speed Controller) is. Obviously a three phase set up, but I'm not sure if it is a sinewave or what. I know it contains a microcontroller and certain parameters can be programmed.

As you are already aware, it's an AC motor (no sensors), the ESC will almost certainly just be switched, no advantage using a sinewave, and LOT'S of disadvantages to doing so - particularly for a model plane!.

 

[Styx]

Of course there is, in as much as you get brushed DC motors, a brushless one simply replaces the commutator with transistor switches triggered at the correct time by hall effect sensors. They both work in an identical fashion - with the main difference being one spins the coils, and one spins the magnet.

*sigh* I mean is as far as there isn't such a thing as a heart attack, it is a MI.
Brushless DC was a phrase conned in a bid to replace the brushed DC
if you examine the flux distrobution in a brushlessDC it has more in common with a sync machine then a brushedDC

I might be just being pedantic at this point

 

[Nigel Goodwin]

*sigh* I mean is as far as there isn't such a thing as a heart attack, it is a MI.
Brushless DC was a phrase conned in a bid to replace the brushed DC
if you examine the flux distrobution in a brushlessDC it has more in common with a sync machine then a brushedDC

I've no interest in examining 'flux distribution', you're getting far too complicated - it simply switches coils in sequence exactly as a brushed motor does.

Your 'no such thing as a brushless DC motor' applies EXACTLY the same to a brushed motor - if you want to look at it that way?, the brushes and commutator convert the DC to AC just like the electronics in a brushless one. And also do the opposite if you use it as a generator, converting the AC from the rotor to DC.

I might be just being pedantic at this point

Yes I think so!.

 

[HiTech]

You're not perhaps thinking of "brushless AC alternators are you? If so the ones that have rotating fields and slip-ring type diodes for regulation produce a clean sinewave output; distortion free enough to safely operate electronic medical equipment.

 

[Hero999]

As far as I'm concerned the diference between a DC brushless motor and a synchronous AC motor, is it has a an electronic comutator i.e. the pulses to the stator depend on the position of the rotor. As far as I'm concerned as soon as you stop using hall effect sensors and start using an inverter it becomes an AC motor.

 

[Nigel Goodwin]

As far as I'm concerned the diference between a DC brushless motor and a synchronous AC motor, is it has a an electronic comutator i.e. the pulses to the stator depend on the position of the rotor. As far as I'm concerned as soon as you stop using hall effect sensors and start using an inverter it becomes an AC motor.

Exactly - and as soon as you remove the brushes and commutator from a brushed DC motor it also becomes an AC motor.