Motor theory debate

We seem to have a lot of really intelegent people on the forum, and every one of you seems to have a unique prospective on things, and a unique field of expertise. So I think this would be the perfect forum for this kind of theroetical question.

To start off, lets say You have two motors, one has few windings, and operates at 24V and will do a max of 20mph. The other has many more windings, uses 48V, and will also do a max of 20mph. Which is more efficent in terms of watts used?

To rephrase it, whats better, low voltage, low winding, or high voltage, high winding motors for efficancy, with power output and speed being the same?

The two motors can be designed diffrently, as long a, for this argument, they are both brushless, and they both have the same power OUTPUT at 20mph, with 20mph being the max speed they can obtain at there respective voltage without any limiter. For ease of explination, both motors are either hub motors, or direct drive, with no difirential gearing.

Now, for a bonous question, which of these two motors would be most efficent at 50% of max speed?

Resistance losses in the copper are the main source if inefficiency.

All other things being equal (power level, size), the 48v setup would be able to run at half the current. The 48v version would have twice the resistance to give the same power level in the same size case. Resistance losses are I2R (current squared times resistance), so the 48v setup would have half the resistance losses. Controller FET and resistance losses would also be 1/4 at 48v.

Switching losses and core losses (fairly small %) would be about the same.

More turns and more volts to make the same power would be the winner. The Ebikes.ca hub motor simulator shows a clear advantage for the 5304 and 5305 vs. a 5303 hub.
If a suitable geared hub doesn't materialize, my next project will be powered by a 5304 in a 20" wheel with up to 150v.

fechter said:

The 48v version would have twice the resistance to give the same power level in the same size case.

Click to expand...

Not twice but quadruple in my opinion.
But if you are twice sure, so tell us detaily why.

I'm assuming power level and physical size are the same for both cases.

To have a 48v run at the same speed and power as the 24v case, the number of winding turns would need to be double. In order to fit twice the turns into the motor slots, the wire would need to be smaller. Wire resistance is a function of length and area.

Twice the number of turns will require twice the length. At the same time the wire must have about half the cross-sectional area to fit.

You are correct, this would result in 4 times more resistance.

Therefore, from a motor standpoint, the two cases would be equal.
I stand corrected.

The resistance losses in the rest of the system (wires, controller, batteries) would then be the main source of difference between the two cases.

My experience indicates the higher voltage systems are more efficient for whatever reason.

fechter said:

My experience indicates the higher voltage systems are more efficient for whatever reason.

Click to expand...

What little experiance I have with Brushless motors says the same thing, more volts=better. But I've never been able to prove it with math.

Your above example shows what happens in the same size case, but what happens when the case size, thickness of wire, number of poles and magnets, and timing can each be optimized for efficancy at the voltage level, and the only requirement is they have the same output RPM and power?

fechter said:

Therefore, from a motor standpoint, the two cases would be equal.
I stand corrected.

The resistance losses in the rest of the system (wires, controller, batteries) would then be the main source of difference between the two cases.

My experience indicates the higher voltage systems are more efficient for whatever reason.

Click to expand...

I can agree now.

So lets try to calculate in which way ( at lowest cost) we can improve the efficiency.
I assume we have choose different motors version and paying only for additional components/parts.

I think the best way to improve the motor is build it with as fat wires as possible.
What you think about ? ( Don't care about my opinion/feeling as i'm not a motor expert )

for 20mph i like 24v
because that's just 2 sla.
very $$$ efficient.

Drunkskunk said:

We seem to have a lot of really intelegent people on the forum, and every one of you seems to have a unique prospective on things, and a unique field of expertise. So I think this would be the perfect forum for this kind of theroetical question.

To start off, lets say You have two motors, one has few windings, and operates at 24V and will do a max of 20mph. The other has many more windings, uses 48V, and will also do a max of 20mph. Which is more efficent in terms of watts used?

To rephrase it, whats better, low voltage, low winding, or high voltage, high winding motors for efficancy, with power output and speed being the same?

The two motors can be designed diffrently, as long a, for this argument, they are both brushless, and they both have the same power OUTPUT at 20mph, with 20mph being the max speed they can obtain at there respective voltage without any limiter. For ease of explination, both motors are either hub motors, or direct drive, with no difirential gearing.

Now, for a bonous question, which of these two motors would be most efficent at 50% of max speed?

Click to expand...

For each of 100s of stator designs there are 100s of rotor designs for BLDC motors.Then you have all the 1000s of otpions for the controllers that can be built to have the same performance by design of each design. Of corse size and weight ratio will be different and so will costs and efficiency as well as a long lasting product... The first thing no one will agree on is the IDEAL POWER and EFFICIENCY RANGE in rpm, torque requirement? , and WATTs to the road requirement ? the efficiency figures can be sorted out.This is a flame war waiting to happen. :lol:

EbikeMaui said:

This is a flame war waiting to happen. :lol:

Click to expand...

Let's not Randy.

I see you've pretty much trashed another thread, and that kind of posting is pointless, counterproductive, and very annoying.
Consider yourself warned. If that kind of behavior continues, I will have no choice but to ban you from yet another bike forum.

I was hoping you would contribute in a constructive manner if we were nice to you, but that seemed to wear off fairly quickly.

We are here to learn and share, not piss people off.

*off soapbox*

There is much to learn about motor design. Current designs are not perfect and there is much room for improvement.

There are many design objectives, which vary depending on application.

Most motor designs are scaleable, so you can scale them to the desired power requirement. For comparisons, assume equal power ratings.

Economics need to be considered as well. A super efficient motor that's too expensive to be practical is no good.

To gain efficiency, look for where the losses are and attack them.
You need to consider the overall system efficiency, not just the motor alone. Batteries, controller, and drivetrain must be considered.

Copper loss is by far the greatest factor for most vehicle applications (except for maybe Peukert effect for us lead users).

Back on the old Voltage Forum, there was a guy named Honk that was building a super efficient coreless axial multi-disk motor from scratch. That was a good design. Too bad I lost track of him. :?

Here's some interesting related stuff:
http://mag-net.ee.umist.ac.uk/reports/P11/p11.html

There is a wealth of good information on that page. I think I need a few IQ points to fully grasp the math used in there calculations, but let me try to summerize what I did understand as applied to this and see if this makes any sense to anyone else?

A higher winding motor is running closer to it's max RPM by design, and therefore is going to have less Cogging Torque loss than a lower wound, lower voltage motor of the same speed and power. End result, higher windings, higher voltage being more efficent.

Thoughts anyone?

Drunkskunk said:

There is a wealth of good information on that page. I think I need a few IQ points to fully grasp the math used in there calculations, but let me try to summerize what I did understand as applied to this and see if this makes any sense to anyone else?

A higher winding motor is running closer to it's max RPM by design, and therefore is going to have less Cogging Torque loss than a lower wound, lower voltage motor of the same speed and power. End result, higher windings, higher voltage being more efficent.

Thoughts anyone?

Click to expand...

Not exactly. A thicker wire with less turns gives a higher RPM at a lower voltage with less resistance.More turns with thinner wire gives you more torque at a lower RPM with a higher voltage but with slightly better efficiency or the same wattage even though the resistance in the winding is higher.At a Higher RMP (within the rated wattage) the thicker low turn windiing will be more efficient as long as the motor stays in its rated wattage range and advoids heavy loads that may not be noticed until heat apears.The thin wire winding would have less peak power in wattage but more efficient for carying a larger load through out the RPM range.Also more copper from thin wire CAN be put on the stator than thick wire. Basically BOTH are the same efficiency. It just depends in how you use each (RPM and LOAD) and how your gear ratio is set up.On most ebikes where the motor often exceeds its RATED WATTAGE the HIGH TURN winding is the most efficient at any speed it can go.. Compared to the low turn thick wire.
Negitive effects of cogging ONLY happen when the power is off and you are coasting.Motors that are connected to a FREEWHEEL will never have any cogging effects.

Motor cogging does detract from efficiency in the form of torque ripple. Torque ripple also causes noise, which is one of the ways the wasted energy gets dissipated. The 'growling' sound common to brushless motors using cheap controllers is caused by torque ripple. Using a trapezoidal or sine wave controller will minimize this.

Usually torque ripple accounts for only a few percent loss at most.

I think the losses from cogging are typically less than core losses.

A freewheel will prevent both losses during coasting, but that means that regen will be impossible. The advantages of freewheeling vs. regen are another topic of debate. In most cases, freewheeling is better (certainly cheaper).

Cogging and ripple current are two different things.Some Motors are wound for sine wave some are NOT.The controller needs to match the motors wave form not the way you mentiond FYI. Core losses happen more above the CONTIONUS RATED MOTOR SPECS.

I am no authority on this but I will agree with fechter and ep's post regarding motor winding resistance and related losses. I did do 2 years full time ed in elec eng but that was 15 years ago!

As regards the efficiency - you really have to look at the total efficiency of the whole system - unless you make a dyno and wish to spend days testing others gear!

IME batteries vary - enormously - I've said this before of course.

Some batteries (esp some nimh) have seriously reduced capacity at discharge rates as low as 1C.

If you do a test with 40 cells in 2p giving 24V on a motor then do same with them in 1p giving 48V (and half the gearing) its not apples with apples - unless you strictly limit current on both tests.....

Scott

scottclarke said:

I am no authority on this but I will agree with fechter and ep's post regarding motor winding resistance and related losses. I did do 2 years full time ed in elec eng but that was 15 years ago!

As regards the efficiency - you really have to look at the total efficiency of the whole system - unless you make a dyno and wish to spend days testing others gear!

IME batteries vary - enormously - I've said this before of course.

Some batteries (esp some nimh) have seriously reduced capacity at discharge rates as low as 1C.

If you do a test with 40 cells in 2p giving 24V on a motor then do same with them in 1p giving 48V (and half the gearing) its not apples with apples - unless you strictly limit current on both tests.....

Scott

Click to expand...

we are comparing motors not batteries.

The original question is what voltage is more efficient.

I agree that you have to consider the entire overall system, including batteries, controller, and drivetrain. That's what counts in the end.

In the case on Nimh batteries, since large cells either don't exist or don't perform well, there is a considerable advantage to running higher voltages.

we are comparing motors not batteries.[/quote]

I know - that being the case real world comparisons not conducted in a controlled manner should not enter into this discussion.

Scott

To get the best results each component need tested individlaly of eacher to find the best of each component to get overall system efficiency.
Example: Why buy lighweight EXPENSIVE batteries and use them with a HEAVY AMP SUCKING motor if there was a alternitive that was lighter and more efficient ? Never mind I know the answer. :lol: People can't make motor mounts.

EbikeMaui said:

To get the best results each component need tested individlaly of eacher to find the best of each component to get overall system efficiency.
Example: Why buy lighweight EXPENSIVE batteries and use them with a HEAVY AMP SUCKING motor if there was a alternitive that was lighter and more efficient ? Never mind I know the answer. :lol: People can't make motor mounts.

Click to expand...

Correct - you got it!

But for those who could work a jigsaw and drill (gasp!) - they may lack experience/confidence in other areas - or may - gasp! - not have a bundle of cash lying around!

I agree non-variable-geared hub motors are less than ideal - especially when climbing mountains.

Personally, I am using a heavy hub motor. You know why? Because they are cheap and I wanted to convert bikes for money. The requirement here is that they are reliable and affordable - at £40 each (+£20 for controller) they fit the bill perfectly!

My geared down outrunner - well the latest gen is costing over £250 including the controller - and thats to deliver 2.2KW at 80%+ eff. via chain and integral hub. Tell me - how much does your 1KW motor/mount/gearing/controller cost?

Scott

scottclarke said:

EbikeMaui said:

To get the best results each component need tested individlaly of eacher to find the best of each component to get overall system efficiency.
Example: Why buy lighweight EXPENSIVE batteries and use them with a HEAVY AMP SUCKING motor if there was a alternitive that was lighter and more efficient ? Never mind I know the answer. :lol: People can't make motor mounts.

Click to expand...

Correct - you got it!

But for those who could work a jigsaw and drill (gasp!) - they may lack experience/confidence in other areas - or may - gasp! - not have a bundle of cash lying around!

I agree non-variable-geared hub motors are less than ideal - especially when climbing mountains.

Personally, I am using a heavy hub motor. You know why? Because they are cheap and I wanted to convert bikes for money. The requirement here is that they are reliable and affordable - at £40 each (+£20 for controller) they fit the bill perfectly!

My geared down outrunner - well the latest gen is costing over £250 including the controller - and thats to deliver 2.2KW at 80%+ eff. via chain and integral hub. Tell me - how much does your 1KW motor/mount/gearing/controller cost?

Scott

Click to expand...

More than yours but I was worth it to have no maintenance except chain lube and a drop of oil once a month for over 20, 000 miles of uninterupted use.