Windings: basic questions.

[learningrc]

Not interested in be "taught" fallacies like: "back-emf is fundamental to how a motor produces power".

A motor doesn't produce power, it converts energy from one form to another, more or less efficiently. And 'back-emf" has nothing, zip, nadda to do with that conversion process; it is the by-product of it.

You are absolutely right ... I shouldn't have written "produces." Completely wrong on my part. I hope you can see, though, that everything I wrote after that sentence referred to "converting" so I would have hoped you could tell it was just the wrong word, not some big "fallacy" or misunderstanding on my part. Kind of like how you say below that the applied voltage "passes though the coils and creates a magnetic field." I'm assuming you know that voltage is just a potential difference and it is current that passes through coils, not voltage.

'Induced voltage' is fundamental to the energy conversion; but 'induced voltage' != 'back-emf'.

For there to be "back-emf" (also termed counter-EMF), there has to be an existing EMF, for it to counter. In a generator, there is no applied voltage, the induced voltage is thus just 'EMF'. Mechanical energy converted to electrical energy by way of magnetism; no "back-emf" required.

In a motor, there is an applied voltage, the EMF. It passes through the coils and creates a magnetic field, that magnetic field interacts with another magnetic field -- the PM in PMSMs -- and that interaction causes mechanical movement. Torque in a rotating motor; just force in a linear motor. And electrical energy has been converted to mechanical energy.

However, a by-product of the movement caused by the interaction of the two magnetic fields is that the relative motion between them induces voltage back into the coils, a voltage that opposes the EMF already flowing there, thus "back-" or "counter"-emf.

EMF (supplied voltage) creates torque; converting electrical energy to mechanical energy.
(In a motor) Back-EMF (induced voltage) is a parasitic by-product of that process, that opposes that conversion and ultimately limits it.

It cannot be avoided completely; but it should be minimised wherever possible to improve efficiency.

Are these the ramblings of a troll?

Trolls are trolls because of their behavior, not because of how knowledgeable they are.

Minimizing back-emf does nothing to improve efficiency. It isn't associated with any losses at all. EI = Tw. If the left hand side of that equation is zero because E is zero, then the right hand side also has to be zero. You literally cannot have power conversion without back-emf.

 

[speedmd]

I am certainly no expert but one thing that stands out and should not be overlooked is that the changing current in the motor coils through induction will create back EMF prior to any motion.

On the Lower turn counts, to my understanding will act more like a short circuit (less able to limit current) than high turn motors. Once the motor starts spinning, the higher turn count motors will induce (magnets passing the coils) a higher voltage at a lower rpm slowing high current flows sooner than the low turn motors. Over all the high turn winding should be easier to control as many here that have stacks of blown controllers can attest.

 

[major]

Throughout this saga I've been expecting someone to step in a support my stance. To say, you guys are conflating your terminology. Browser is obviously correct. That hasn't happened....

Have you considered why?

the only logical conclusion is that one person got the wrong end of the stick and everyone else cargo-culted it.
That is the only explanation for some many people apparently believing that something that, by definition, cannot possible exist until the motor rotates, is responsible for causing it to rotate.

Please indicate what you are talking about with direct quotes.

 

[Browser]

If you use S.I. units of Nm/A for Kt and V/(rad/s) for Ke, Ke=Kt. The proof of this fact is available all over the internet in open University courses, text books, PHD thesis papers, IEEE papers etc. By minimizing Ke , the Bemf constant, all you achieve is to minimize Kt. Bemf and torque production are bound together by the laws of physics. In practice the choice of Kt, Ke are dictated by the specific design requirements of the application. What is the line voltage, what are the power requirements at what rotational speed? Low Ke motors are generally low inductance motors which follows from the fact that inductance is proportional to the number of turns of wire employed. We all know that low inductance motors are pigs to control because low inductances running on commonly employed line voltages lead to exceptionally high di/dt rates that make it imposing the pwm voltage wave required to achieve FOC near impossible with current technology. Since we want the maximum torque per amp delivered by FOC we don't want motors with exceptionally low Ke=Kt.

Yes. Unfortunately it seems they are very easy to misinterpret.

I think the engineers at YASA motors, a spin-off from Oxford University, are some of the smartest electrical engineers on the planet. Take a look at their design, it doesn't appear that they are of the opinion that low Ke is the path to efficiency and I agree entirely.

You mean Tim Woolmer, and Malcom McCulloch formerly of Oxford U. in 2006. (The former now CTO at http://www.yasamotors.com.) Their's is a 10-pole/12-coil, 3600rpm axial flux traction motor. (Ie very low speed compared with mine.)

Despite that, their paper http://www.mojaladja.com/upload/elmotor/Analysis of the Yokeless and Segmented Armature machine.pdf (7 pages with great pictures) spends most of the time considering the eddy current losses.

A few snippets:

The most significant eddy currents in the stator are those induced in the stator iron, since it observes the largest changing flux density. However, the motor casing and copper winding are also affected by eddy currents, but to a lesser degree. The powdered iron material used in the machine has a relatively high electrical resistivity of 400μΩm. The induced eddy currents in the shoes and bars of the machine are shown in Fig. 9. Despite the high resistivity of the SMC (soft magnetic composite) material, the high machine electrical frequency (300Hz) and large area for eddy current paths mean that there are significant losses caused by eddy currents.

D. Rotor Eddy Currents The rotor eddy currents are caused by the interaction between the stationary field of the stator and the movement of the rotor through that field. NdFeB magnets have a conductivity of approximately 16000μΩm, approximately 10 times less than mild steel, which can lead to a significant heating affect of the magnets. Reducing this heating affect is important, not only because the magnets have a maximum working temperature of 180˚C, but also because the energy product of the magnets is significantly reduced as their temperature is increased.

And from their conclusions:

Finite Element analysis has been used to accurately calculate the machine losses for the LIFEcar YASA machine. The results have shown that SMC material produced by Höganäs performs well, although significant eddy currents add to the overall losses at electrical frequencies about 250Hz. One possible technique to reduce these eddy currents is to laminate the SMC segments radically. FEA has also revealed that magnet segmentation and coating are particularly important for controlling the rotor losses. An epoxy coating will significantly reduce rotor losses since it stops eddy current paths between the magnets and back iron.

Then, if you feel motivated to learn more, you might want consider reading ""Analysis of the proximity and skin effects on copper loss in a stator core" http://www.degruyter.com/dg/viewart...sue-2$002faee-2014-0017$002faee-2014-0017.xml

A few snippets:

There are two principal mechanisms for losses in high speed ac motors at high frequency, that are iron losses and copper losses. The most significant losses come from the winding due to copper resistance. High speed motors at igh-frequency sinusoidal current waveform, which can make the effective winding resistance very high and for that reason copper losses are increasing when frequency grows [1, 2]. The winding resistance of high speed motor is a result of both skin and proximity effects [2-5]. The main phenomenon that leads to ac losses is the skin effect which is the tendency for high-frequency currents to flow on the surface of a conductor. The skin effect can be mitigated through the use of smaller conductor strands.

Alternating currents at high frequency are confined to a thin layer at the surface of the con-ductor due to the skin effect. The general theory shows that the current penetrates a depth δ that can be expressed: {skin depth formula here}
,
The second phenomenon is the proximity effect which is the tendency for current to flow in other undesirable patterns that form localised current loops or concentrated distributions due to the presence of a magnetic field generated by nearby conductors. The proximity effect can be mitigated by transposing conductor strands. Both of these are the well-known current-displacement effects within electrical machines, transformers and inductors.

During high rotating flux at high frequencies, the ac resistance can be prominent, usually exceeding dc resistance and resulting in high copper losses.

(S'called research!)

 

[Miles]

(S'called research!)

Somehow, I don't think any of this is news to the contributors to this thread..... It's also completely irrelevant to the disagreement we are dealing with.

If you could turn down the patronisation a notch or two, it would be appreciated.

 

[Browser]

If you could turn down the patronisation a notch or two, it would be appreciated.

Really. How about "Somehow, I don't think any of this is news to the contributors to this thread". Patronising much?

 

[Browser]

It's also completely irrelevant to the disagreement we are dealing with.

Oh And what disagreement is that? (And how come you get to decide what my thread is about?)

 

[Miles]

Oh And what disagreement is that? (And how come you get to decide what my thread is about?)

That the back EMF constant, itself, has anything to do with efficiency.

It's not "your" thread. It's a discussion which you started.

 

[Browser]

That the back EMF constant, itself, has anything to do with efficiency.

S'funny, but I've never mentioned "back-EMF constant".

It's not "your" thread. It's a discussion which you started.

Really. So I come to ask about one thing, and you're going to ram your alternative choice of topic down my throat. Now who's the troll?

 

[Punx0r]

To clarify: You believe BEMF to *only* be the result of combined eddy currents induced within the winding conductors and is an entirely parasitic loss?

If so, it would explain your desire to minimise BEMF. Although it leaves me confused what would happen to the current draw and efficiency of the machine: At no-load, wouldn’t it be drawing the same current as at stall, but almost all the power being absorbed as ohmic heating in the winding due to huge eddy currents?

 

[Browser]

To clarify: You believe BEMF to *only* be the result of combined eddy currents induced within the winding conductors and is an entirely parasitic loss?

Close enough. Yes.

Although it leaves me confused what would happen to the current draw and efficiency of the machine: At no-load, wouldn’t it be drawing the same current as at stall, but almost all the power being absorbed as ohmic heating in the winding due to huge eddy currents?

The emboldened bit is a little confusing as you don't say under what conditions you think this would occur; but I'll assume you mean if I could completely eliminate backEMF?

If so, then as I told lebowski, I've never for one moment suggested that I want to; much less could. It is impossible to eliminate it; but it can be minimised.

Stepping back a little. The moving magnets in PMSM induce voltages in any impinged conductor. The familiar example is iron cores, where the induced voltage is normally termed "eddy currents", universally recognised as a parasitic drain upon efficiency and routinely minimised through laminations or using soft magnetic composites.

But when those same moving magnets impinge upon the coil conductors, induces voltages are termed 'backEMF'. Same process; moving magnetic fields interacting with conductive material.

Except, unlike the electrically isolated cores where all that induced voltage can do is circulated in local areas, drive current through resistance and produce heat, the voltages induced in the coils are in circuits carrying an existing EMF of opposite polarity. And a smaller current cannot flow against a larger current (water analogy), so its net effect is to reduce but not eliminate that opposing flow. Until such time as the motor speed increases to the point where the speed of the changing magnetic field induces a voltage equal to, but opposite to the EMF flowing, at which point, a net no current is flowing, so no more torque can be produced. Max speed is reached.

If you can eliminate some of the backEMF, then you raise the maximum speed of the motor; but, more importantly, that reduction in backEMF, means more of the supplied EMF makes it to the coils where it can create a stronger magnetic field, thus more peak torque, and peak power; but also more power for the given supplied EMF throughout the entire rev range, and that equates to greater efficiency.

Of course, it is a balancing act. You can go to more and more turns of finer and finer wire, but you increase the basic DC resistance, with all that that entails.

Or you go the other way; use a single, big conductor for each coil -- see the pdf I linked back on page one or two -- which decreases your DC resistance losses, but the large cross-section (as presented normal to the magnetic field--think lamination direction) means that you get huge induced voltage: backEMF, which limits top speed, maximum torque and torque/given input EMF setting throughout the range.

So, why not look for a compromise. You want minimal cross-section presented to the changing magnetic field, without reducing the cross-section of the conductors to the point where DC resistance goes through the roof. So, why not use foil-wound coils. Arrange for the thin edge to be presented normal to the magnetic field, whilst the wide face ensures that the conductors aren't so long and thin that they create huge DC losses.

This is not a new idea. In 1990, the Uni. of Wisconsin produced a design for "A low loss Permanent Magnet Brushless DC motor utilising Tape-Wound amorphous iron" http://lipo.ece.wisc.edu/1990pubs/90-18.pdf.
(I wish I'd found that paper before I came here asking my questions; maybe all this would have been avoided, but you can't change history

 

[Miles]

Thanks for providing that excellent summary of your misconception!

 

[Browser]

Thanks for providing that excellent summary of your misconception!

And thank you for ignoring the logic, evidence, and basic physics and providing your unwanted, unwarranted, unhelpful, unnecessary, not to mention smug, snide and patronising commentary.

 

[Miles]

Thanks for providing that excellent summary of your misconception!

And thank you for ignoring the logic, evidence, and basic physics and providing your unwanted, unwarranted, unhelpful, unnecessary, not to mention smug, snide and patronising commentary.

My pleasure. Sorry for not disturbing your reverie.

 

[Browser]

And now for something completely different.

A question (assuming you built it yourself?): What are the motor constants for your eMoulton machine? And how did you ascertain them?

 

[Miles]

And now for something completely different.

A question (assuming you built it yourself?): What are the motor constants for your eMoulton machine? And how did you ascertain them?

No, I didn't build that motor. It was one of a small batch put together by South West Windpower http://www.irecusa.org/2013/05/what-happened-to-southwest-wind-power/

https://endless-sphere.com/forums/viewtopic.php?p=79670#p79670

 

[Browser]

No, I didn't build that motor. It was one of a small batch put together by South West Windpower http://www.irecusa.org/2013/05/what-happened-to-southwest-wind-power/

https://endless-sphere.com/forums/viewtopic.php?p=79670#p79670

Oh. Okay.

(BTW: the pdf on the linked page https://endless-sphere.com/forums/download/file.php?id=11514 appears to be corrupted?)

 

[Miles]

Yes. Unfortunately, a lot of the attachments on the ES server were corrupted by an update. Anyway, it's just a model of the case IIRC.

The motor isn't anything special really....

 

[Browser]

And another question.

Ignoring the horrible impracticality -- and expense -- what would happen if you put zener diodes between each turn in all of the coils in your motor so that they allowed the supply voltage to flow, but inhibited the backEMF?

 

[major]

This is not a new idea. In 1990, the Uni. of Wisconsin produced a design for "A low loss Permanent Magnet Brushless DC motor utilising Tape-Wound amorphous iron" http://lipo.ece.wisc.edu/1990pubs/90-18.pdf.
(I wish I'd found that paper before I came here asking my questions; maybe all this would have been avoided, but you can't change history

Interesting paper by Tom. However he uses round standard AWG size copper wire for the coils. BTW, I invented that motor five years earlier.

 

[Browser]

I invented that motor five years earlier.

Really!?

 

[major]

Really!?

Everything I have written and posted on this thread, or on this forum is true.

 

[Browser]

Everything I have written and posted on this thread, or on this forum is true.

Sorry. There was no implied accusation.

Just, what are the odds of me referencing half a dozen of the hundreds of papers on the subject that I've read or scanned and the thousands of other out there,and one of them hits home amongst a very small, even if self-selecting, group of participants in a single thread of a forum.

 

[Punx0r]

The internet is an amazing resource partly because it is often possible to gain valuable advice from experts, who give their time freely, which out in the real commercial world you would have to pay very handsomely for - if you even knew where to find it.

I think we sometimes forget how hard it was to find specialist knowledge before wide participation in the WWW and for the youngest amongst us they have never known it...

 

[Punx0r]

Oh, and care to draw a circuit diagram or two showing voltages and current for your diode idea in its on/off states? It's a little hard to picture and I have a feeling quantifying the variables might provide some clarity.