Brushless motors-how many poles? / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Brushless motors-how many poles?

previous 1 2

Move Thread

LAN_403

Ash Small

Tue Aug 05 2014, 02:58PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Steve Conner wrote ...

I can't argue with that, but I also find it non-intuitive that the answer is independent of the number of poles in the stator winding.

The article said that due to a different winding configuration, the stator doesn't generate a rotating magnetic field like in a classical polyphase motor. It is a pulsating magnetic field and the 3 pole pairs act like 3 independent single phase machines. So adding pole pairs (triplets, sextuplets, whatever) to the stator doesn't change the speed, it just increases the torque.

If you compare it to a stepper motor, which is effectively two phase, and, depending on which 'pair' changes determines the direction that it 'cogs', ie 1st phase positive, 2nd phase positive, 1st phase negative second phase negative produces 4 'cogs' in one direction, which moves one stator pole to the next magnet of the same polarity.

In the three phase system, six 'cogging events' are required to move one stator pole onto the next magnet of the same polarity. There are 7 magnets of each polarity in your motor, which requires 42 cogging events (7 x 6) for one complete rotation. There are 6 cogging events per three phase cycle, so it takes 7 cycles (42/6) for one complete rotation.

BigBad

Tue Aug 05 2014, 05:07PM

Registered Member #2529 Joined: Thu Dec 10 2009, 02:43AM
Location:
Posts: 600

No, it is actually a rotating field; in the animated diagram they're only showing you the moments when the fields line up for simplicity.

It says:

"Praxis:

The 3-phase field is rotating at 42,000 rpm, the magnets bell and he propeller at 42,000/7=6,000rpm."

They've also got diagrams of the driving voltages above, and they're fairly normal (non sinusoidal) 3-phase waveforms. It's not pulsed.

edit: cool related Moire gif:

Movement

Notice how much faster the banded pattern moves; that's the drive frequency.

Steve Conner

Tue Aug 05 2014, 05:09PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

I'm not trying to "design a motor starting from hints", I'm trying to understand the general principle behind the "outrunner" brushless motors that have recently become very popular for quadcopters and the like.

I am more or less satisfied by Dr. Slack's answer, that the number of stator poles doesn't figure in the equation, because the motor simply won't work unless it bears the correct relation to the number of rotor poles. However, the literature on outrunner motors says that a few other ratios work besides 14:12.

BigBad

Tue Aug 05 2014, 05:22PM

Registered Member #2529 Joined: Thu Dec 10 2009, 02:43AM
Location:
Posts: 600

It's basically just sampling theory.

Any ratio of poles should work pretty well to reduce cogging provided they have a coprime relationship; in this case 7:6 because at any instant that averages out the different relative positions of the stator poles to the magnet poles around the circle.

Also: of course more poles are better, less poles are worse.

Steve Conner

Tue Aug 05 2014, 05:50PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

OK, I think I get it. In the particular case of the motor with 14 magnets and 12 coils: The stator field rotates at 7f, but it is sampled spatially in 12 places, so the Nyquist frequency is 6f, and there will be an "aliased" field rotating at 1f, which is the speed of the rotor.

(Edit: Maybe I meant 6 places rather than 12, but either way the argument seems reasonable.)

I just need to prove that the rotor magnets will follow this particular alias frequency in preference to the others. I guess by the principle of least action, you could show that the lowest alias frequency will generate the greatest torque.

Ash Small

Tue Aug 05 2014, 06:02PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

BigBad wrote ...

They've also got diagrams of the driving voltages above, and they're fairly normal (non sinusoidal) 3-phase waveforms. It's not pulsed.

It's exactly the same principle as stepper motors, except this has three phases instead of two (to reduce the cogging effect, which is utilised in stepper motors) and is driven with a pretty constant waveform (apart from speed control). The last stepper motor I took apart had 48 poles and 50 magnets, if I recall correctly.

radiotech

Sun Aug 31 2014, 10:33PM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

My answer to the original question:

My question is, since the rotor and stator have different pole numbers, which one do you use to calculate the speed? Put another way, if I removed 2 magnets from the above motor (and redistributed the remaining ones evenly) would the synchronous speed stay the same, or would it increase by a factor of 14/12?

Is this:

In a poly phase induction motor, the synchronous speed is f * (n / 2) f= Hz/sec, n = number of poles on the stator This yields revolutions/ second. A 2 pole machine on 50 Hz. runs at 50 revs/second or 3000 revs per min.

The speed is a function of the number of poles on the stator.

This has been the case ever since Tesla made the first practical one.

Steve Conner

Mon Sept 01 2014, 07:27AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

radiotech

Tue Sept 02 2014, 05:21PM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

Way to go Radiotech, you just missed the entire point of the thread. The example motor we are using, with a 12 pole stator and 14 pole rotor, spins at one-seventh the speed predicted by that equation.

I agree with missing the point and later got this information from a book that I used
when stepper motors was a concern. And the page of interest was marked !.

m is number of phases.

For the case where the teeth(n) on stator exceeds rotor, and vice versa, the steps
per revolution(S) is different:

previous 1 2

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob