# Same debate again, motor windings!

wiki says cogging torque's effects are filtered out by inertia at higher speed but it doesn't say any less power is needed to overcome cogging torque and I still believe it's rpm dependent. it may be smoothened by inertia but the loss is still there no?

where does the momentum go when lost to cogging torque? what kind of loss would it be categorized as? if you get a motor with really bad cogging torque it wont coast far at all and where is all that momentum going?

wiki says cogging torque's effects are filtered out by inertia at higher speed but it doesn't say any less power is needed to overcome cogging torque and I still believe it's rpm dependent.

The force may be constant at different speeds, but the power wouldn't be constant. In that case it would be proportional to speed (P=FxV).

where does the momentum go when lost to cogging torque? what kind of loss would it be categorized as? if you get a motor with really bad cogging torque it wont coast far at all and where is all that momentum going?

You've got a very small fluctuation of a magnetic field through a magnetic material. So the losses are mechanical (friction, noise, vibration), electrical/magnetic (eddy currents and hysteresis). All of these things end up as heat, one way or another.

Now I'm on a mission to find Justin's post I remembered.

In the meantime here's one pretty close. Cogging doesn't net to zero, and the post kinda says that it sort of plateaus at low speed. I guess that sponge between my ears is still functioning ok.

justin_le said:
dogman said:
I find cogging varies with speed. I feel very little at 5 mph, more at 10 and so on.

The actual cogging torque of most direct drive hubs if you care to measure it does increase with speed but not by very much (not enough for you to feel the difference, it might be 0.4 N-m at 5mph, and like 0.45 N-m at 10 Mph).

Justin

John in CR said:
Now I'm on a mission to find Justin's post I remembered.

In the meantime here's one pretty close. Cogging doesn't net to zero, and the post kinda says that it sort of plateaus at low speed. I guess that sponge between my ears is still functioning ok.

justin_le said:
dogman said:
I find cogging varies with speed. I feel very little at 5 mph, more at 10 and so on.

The actual cogging torque of most direct drive hubs if you care to measure it does increase with speed but not by very much (not enough for you to feel the difference, it might be 0.4 N-m at 5mph, and like 0.45 N-m at 10 Mph).

Justin

It's funny, but both of those comments seem like different things, but actually both can be true.

As Justin says, the cogging torque remains relatively constant at different speeds. That makes sense.

Dogman says he thinks it varies with speed. Which can be true also. Because if the torque remains constant, the power will vary in proportion with speed, and the power is the energy exertion over time that a human feels (in this case torque is synonymous with force, and power = force x velocity).

So the drag force produced by cogging is a constant (based on the motor geometry), and the power required to overcome that drag force rises linearly with speed.

I doubt anyone feels power, i think torque onset and acceleration is what we can subjectively feel.

A premium car going at 100mph on the freeway is outputting a lot of power, yet you think it’s going only 50 mph compared to the feel of that old nissan your grandma drives.

We are very poor at sensing stuff, and i am sure i will never meet a guy that can sense 0.05Nm of torque difference on a hub. That would be a golden ass :wink:

the torque ripple is smoothened by the inertia and you won’t feel it but it’s a loss still. Less range especially if u coast.

But where does the momentum go that gets lost to cogging? A slotless motor in theory should have greater iron losses than a slotted motor as there’s more conductive material passing permanent magnets, but slotless motors have almost no cogging. Seems cogging is not an iron loss.

larsb said:
I doubt anyone feels power, i think torque onset and acceleration is what we can subjectively feel.

A premium car going at 100mph on the freeway is outputting a lot of power, yet you think it’s going only 50 mph compared to the feel of that old nissan your grandma drives.

We are very poor at sensing stuff, and i am sure i will never meet a guy that can sense 0.05Nm of torque difference on a hub. That would be a golden ass :wink:

I think Dogman was talking about pedaling the motor unpowered, and with some of the old thick lam high pole count hubbies back then that the other iron losses were kicking in enough to feel by 10mph. It doesn't take much drag to feel with your legs while unpowered. Maybe it was an old angled stator CLyte that doesn't really have cogging.

My point about Justin's quote, was that he guesstimated a cogging increase of only 12.5% with a doubling of rpm from 5 to 10mph, ie kinda plateauing at a relatively low level. I wish we could get him in here to clear things up, as he's measured all this stuff.

I think this shows a three phase motor with roughly 1/50 of the torque being lost to cogging during rated speed. That’s while powered and I imagine the cogging torque will be maybe 2x higher when unpowered when all teeth will be pulling at magnets instead of possibly pushing them with electromagnets)

The cogging torque is shown to stay the same at double speed.

permaloy in the slots over the winding to greatly reduce cogging:

John in CR said:
I think Dogman was talking about pedaling the motor unpowered,

I agree. That's what I was referring to also.

You're not going to feel cogging torque in a powered motor. Other magnetic forces far outweigh that effect when under power.

Regardless of if it’s seen as insignificant when powered or unpowered..what is it? How does the wheel momentum disappear and does the cogging torque loss increase linearly w speed?

And would the cogging torque be half when powered vs coasting?

Regardless of if it’s seen as insignificant when powered or unpowered..what is it? How does the wheel momentum disappear and does the cogging torque loss increase linearly w speed?

This is like talking to a brick wall. SMH.

still haven’t gotten an answer as to what cogging torque is or where the momentum goes with lots of searching and not posted here. I don’t see how it would be eddies n hysteresis as a slotless motor has no cogging yet more conductive material passing magnets. Judging by the posts above it’s still unclear if cogging loss even increases linearly with speed. You bring generalities with no evidence and make it out like I’m just not getting it. Evidence of anything?

I too don't understand what you mean by "How does the wheel momentum disappear?" . If it's what causes the wheel to slow down more than an unmotored wheel, eg when a DD hubmotor is spun in free air by hand in addition to wind resistance of the wheel, tire and spokes, and bearing friction, then it's (all of which turn into heat in some manner):
1. Cogging torque
2. Eddy current torque
3. Hysteresis torque
4. a tiny bit of windage (wind resistance) inside the motor.

Sure. And how would cogging torque transfer to heat?

I’ve been looking down this road before and remember cogging did not produce heat. Seemed more magnetic friction which doesn’t seem to have a loss

You've got a very small fluctuation of a magnetic field through a magnetic material. So the losses are mechanical (friction, noise, vibration), electrical/magnetic (eddy currents and hysteresis). All of these things end up as heat, one way or another.

Wikipedia, eddy currents:
Eddy currents are loops of electrical current induced within conductors by a changing magnetic field in the conductor, current through a conductor creates heat through resistance.

Wikipedia, Induction heating:
Magnetic materials improve the induction heat process because of hysteresis. Materials with high permeability (100–500) are easier to heat with induction heating. Hysteresis heating occurs below the Curie temperature, where materials retain their magnetic properties.

https://www.motioncontroltips.com/hysteresis-loss/
Hysteresis loss is caused by the magnetization and demagnetization of the core as current flows in the forward and reverse directions.

= wasted energy from cogging becomes heat.

In the study I linked above they used permalloy over top of the windings, filling the slot, to greatly reduce cogging torque. This is adding more conductive and ferromagnetic material in the path of moving permanent magnets to DECREASE cogging.

Another method of reducing cogging is decreasing the slot width thereby also adding more material in the magnets’ path. Skewing teeth or magnets is another way which also doesn’t make sense in reducing iron losses yet reduces cogging a lot.
Lately thinking it’s “magnetic friction”.

Filling in the gap above the windings reduces the variation in the ferromagnetic material from pole to pole, and therefore reduces the amplitude of the magnetic fluctuations.

Skewing the stator laminations spreads each "cog" incident over a greater distance where each "cog" overlaps the next, reducing the amplitude of the fluctuations (analogy: single phase vs multu-phase, and how there is less ripple in multi-phase power).

the mellow hub motor here: https://www.electric-skateboard.builders/t/mellow-production-stator/10231
it has no slots between teeth. It has almost no cogging torque. If iron losses were the cause of cogging..these teeth are all exposed to the same field strength just with no gaps between and should experience roughly the same iron losses as if it had slots. With no slots there’s a coast as easy as a regular bearing. Adding 1.5mm slots between teeth isn’t going to produce lots of hysteresis especially at very slow speed

You're getting iron losses caused by the permanent magnets alone (unpowered motor), confused with combined iron losses caused by the coils and permanent magnets (powered motor).

ur saying I’m confusing powered Vs unpowered.. the iron losses will be increasing when the motor is powered as more of the stator will be exposed to a switching field and the field is stronger, but cogging torque/ripple is greatly reduced when the motor is powered.

I’m looking for evidence not off top of head. The example of the mellow motor shows me cogging torque is not caused by iron losses as it has ferromagnetic material wrapping the whole stator diameter exposed to the magnets’ field yet no cogging. I think all the material on the diameter of stator will be exposed to the same switching field regardless of the slots or lack of slots. Seems safe to say we’re talking about the pull of a magnet from iron not hysteresis which is minuscule at low speeds

what is cogging torque and where does the momentum go? Maybe nothing is heated by cogging torque and if I go down a mountain coasting it’s like friction without heat produced...but that seems to conflict w conservation of energy

Who’s gunna argue with me? Huh?! Really trying to get to understand it. Not even the thread

https://www.electric-skateboard.builders/t/freewheel-spinning-on-a-mellow-drive/14497

Cogging torque is a "sine save function (repels and attracts equally)" - Mellow boards themselves

Just picture a single magnet spinning past some stator teeth. When it is aligned with a tooth, forces are balanced. As you rotate the magnet away from that tooth towards the next tooth, the initial tooth pulls on the magnet with an initially strong force, but it gets weaker and weaker as the magnet moves away. The next tooth pulls with an initially weak force, but it gets stronger and stronger as the magnet moves closer to it. Halfway between the teeth, forces are again balanced. As you keep moving the magnet, the next tooth pulls harder than the initial tooth, and all the energy you expended to pull it away from the first tooth is given back to you as the next tooth pulls it towards it.

Work is force times distance, force varies sinusoidally-ish with distance. The important thing is that the force alternates between positive and negative symmetrically. Thus the integral of force over distance is 0 for 360 electrical degrees.

If you give a motor a spin and let it coast to a stop, at some point the kinetic energy of the rotor is not enough to overcome the energy required to pull away from one tooth to the halfway point between teeth, and the motor falls back towards the previous tooth. It oscillates around the tooth a little bit, but it is also losing energy to hysteresis losses and eddy currents, so it doesn't oscillate long.

The mellow hub does not freewheel - it still has hysteresis and eddy current losses - it just spins a little more smoothly at low rpm since the peaks and valleys of the above graph are much smaller. Smaller peaks and valleys mean the minimum kinetic energy of the rotor to keep spinning is lower, and the effect of pulling away and toward each tooth is smaller.

thepronghorn,

It seems like you're saying that cogging nets to 0, and I don't believe that's true. A free spinning DD hubmotor simply doesn't slow at low rpm in that manner. If cogging netted to 0, then during a spin down it would only come into play as the wheel stops with the last "notch". It would also mean that a spin down would happen in a similar fashion to a simple wheel and bearing, just slowing more quickly due to the additional losses that increase with rpm. During the last revolution or two, just like with bearing friction loss, the rpm related losses hysteresis and eddy currents become so small that momentum would make the last revolution or two take a while. Instead just the opposite happens, and a DD hub slows much faster in the last turn or 2 before it stops, which indicates there is a force closer to constant that the spinning motor must overcome. I call that cogging.

I've never seen anything showing the pull force of a magnet would be different if joining or separating but assuming that's the case wouldn't the cogging be sum zero beyond a minimum input of kinetic energy? as said. kind of like this scooter. imagine you could make a magnetic model similar to this scooter with one very rounded tooth and magnet.

still think eddies and hysteresis don’t seem the cause of cogging when looking at its resistance produced at different speeds.

what other possibilities are there?

#### Attachments

• 3CE73CE1-953D-4EBC-8616-1398B21BDEBC.png
5.7 KB · Views: 1,038
Straight from Wikipedia

Cogging torque of electrical motors is the torque due to the interaction between the permanent magnets of the rotor and the stator slots of a Permanent Magnet (PM) machine. It is also known as detent or 'no-current' torque. This torque is position dependent and its periodicity per revolution depends on the number of magnetic poles and the number of teeth on the stator. Cogging torque is an undesirable component for the operation of such a motor. It is especially prominent at lower speeds, with the symptom of jerkiness. Cogging torque results in torque as well as speed ripple; however, at high speed the motor moment of inertia filters out the effect of cogging torque.

Just because we only can feel the vibration at very very low rpm doesn't mean it doesn't occur up higher. That's the answer to your "where does the momentum go", as others have stated above regarding vibration, and those vibrations, or micro-vibrations if you will, turn into sound and heat. There is definitely a torque drag from cogging, so it doesn't net to zero. It also doesn't continue climbing with rpm like the other iron core losses.

Replies
9
Views
1,428
Replies
2
Views
145
Replies
3
Views
304
Replies
2
Views
147
Replies
4
Views
256