Geared hub are 15-25% more efficient than DD

[cwah]

Hello there,

I've been on Ebike-kit site, and I've seen there table comparison between geared hub and direct drive:
350w Internally Geared Motor
Advantages vs. Direct-Drive:
Lighter Weight (7 pounds vs. 14 pounds)
Smaller (134mm diameter vs. 244mm diameter)
Zero Drag (complete freewheel when not using battery)
Excellent Efficiency (15-25% more range)

Disadvantages vs. Direct-Drive:
Lower Top Speed 36v (17.5 MPH) / 48v (24 MPH)
Less Power (350w vs. 500w = 40% hill-climbing power)
Moderate Life span (internally geared parts wear down. 2+yrs expected)
*Internal gears are replaceable by E-BikeKit

500w Direct-Drive Motor
Advantages vs. Geared:
High Top Speed 36v (20 MPH) / 48v (28 MPH)
More Power (500w vs. 350w = 65% hill-climbing power)
Excellent Life span (brushless and no friction = "last forever" 5+yrs expected)

Disadvantages vs. Geared:

Heavier Weight (14 pounds vs. 7 pounds)
Larger (244mm diameter vs. 134mm diameter)
Slight Magnetic Drag (slight resistance when not using battery)
Moderate Efficiency (15-25% less range)
http://e-bikekit.com/electric-bike-conversion-kit-motor-for-bike-e-bikekit.html

Direct drive would have 15 to 25% less range under the same condition? that's huge.

Do you have the same result from Direct drive hub?

 

[dogman dan]

No. Emphatically NO.

When I was doing a lot of motor testing and review, I found that if you ran a small gearmotor at 20mph, it used about 400w. And if you ran a larger dd hubmotor at 20 mph, it used about 400w. It was so close, it was nearly identical.

HMM, maybe I need to do a longer test? Ok, ride the two motors, on flat ground, in identical good weather, till the same battery reaches its cutoff.
Both motors ran about the same distance, 28-29 miles. Add some hills to the test, and the results are about the same, shorter range with the hill, but if speeds are identical still, the range is very close to identical.

So why do people get such different results? They have different riding conditions. Or you have changed a second variable. It might be hard to ride two motors up steep hills the same speed.

The gearmotor did much better efficiency if the ride had lots and lots of stop signs. The better take off makes less heat than a typical dd motor. However, if you limit speed to 20 mph, then using a slow winding dd motor gets about the same results. It's why I have a slow winding dd on my big heavy longtail.

The gearmotor gets much much much better efficiency if you ride it slower. Wot riding on a 20 mph setup will be much more efficient than riding wot on a setup that does 27 mph.

The gearmotor, or any motor for that matter, will go much much farther if you pedal with the motor off. If you want to do pulse and glide riding, then you want a motor that freewheels.

But if you ride like I do, at about 20 mph, on a route that is long and has stop signs only every few miles, and are motoring all the time, there is not enough difference to measure. If you have downtown city riding, you want a lower speed winding, or a gearmotor.

One last thing, I didn't like the small gearmotors near as much because sometimes I needed more than 400w. I have a big steep hill, but what really turned me off about the tiny motor was a big headwind. With only 400w, I could find myself facing miles and miles of riding no faster than 14mph, because I had a 30 mph headwind. Nicer to have 1000w, even when you mostly still ride along using only 400.

Because a larger gearmotor does not have this problem, I tend to point people to the 8t Mac a lot now. It has the good take off at a stop sign, yet isn't so limited in top speed or wattage up a steep hill as the tiny gearmotors.

But the tiny gearmotors do have thier place. Ideal for a lighter roadbike, and a rig designed to get up just one steep hill along the way, using a battery that weighs only 3 pounds. Or a rig where lower speeds and shorter distances are what you want. Nice for a beach cruiser intended to ride on the beachfront bike trails at 10 mph.

 

[dogman dan]

One last comment. Though I found efficeincy identical when you eliminate the other variables, in the real world you likely won't ride that way.

Give me 350 watts and I can't use 750w. Give me 750w, and likely I'll climb hills faster, and may accelerate faster, or simply ride faster. For sure I'll ride faster into a hard headwind. So in the end, likely I will use more wh/mi on the big dd motor.

The difference is not so much geared vs dd, it's just 350w vs 750w.

 

[SamTexas]

Here's my personal experience:
http://endless-sphere.com/forums/viewtopic.php?f=30&t=34998

Dec 28, 2011

I have a love/hate relationship with my 9c 2807 DD hubs. I like them because they have no mechanical parts and because they can take more power. I hate them because they are so inefficient. I have done many tests comparing the 9c 2807 with the cell_man MXUS 350w geared kit. Running on the same test loop, on the same bicycle, maintaining the same riding style and average speed, the 9c 2807 always consumes AT LEAST 20% more energy.

I would like to know exactly (or as exactly as possible) the source of that inefficiency and its magnitude at various speeds. It appears to me that the higher the speed, the less efficient it gets. What kind of test, what kind of instruments do I need to get solid, repeatable scientific numbers?

Thanks.

I'm still looking for a scientific explanation for the inefficiency of DD hubs (compared to geared).

 

[o00scorpion00o]

On level ground there probably won't be much difference , but the geared hub will defiantly have more torque for the hills than the direct drive going up faster with the same voltage and current, so making it more efficient.

Some might argue that if using a direct drive you can use regen on steep hills giving some efficiency back. But you won't generate much unless going down a very long hill !

one of the differences between geared and direct drive is the power the geared motor will deliver compared to a direct drive of the same size, the geared hub will have a lot more torque.

Probably the biggest thing is the geared hub will be much easier to pedal compared to a direct drive and anyone that likes to pedal at all won't like the drag of a direct drive motor, the free wheeling and ease of peddling is why I love geared hubs best!

 

[Philistine]

My subjective experience (from memory, no records) of comparing a BMCV3, and a 9C 8x8 as well as a X5303, was that the BMC was definately more efficient (maybe 10-15% at least).

I'm still looking for a scientific explanation for the inefficiency of DD hubs (compared to geared).

I would have thought that would make scientific sense, because a geared motor has a greater ability to stay closer to its efficiency point, being able to potentially rotate faster than the axle. I would have thought it would make sense in the same way it makes sense that a middrive through the gears would be more efficient than a DD (not in terms of the difference in efficiency, but more efficient for the same reason, ie, the motor can turn at a faster speed than the axle).

 

[Drunkskunk]

they are compairing apples to oranges. I happen to have both kits from them. I actualy get a little further on my direct drive kit.

But they are comparing a small motor to a large motor, of two diffrent technologies. Its like saying a 40 watt lightbulb will save more power than a 75 watt bulb. True, but that ignores the reason for having a bulb. Light. and the 75 watt will give more light. a 350 watt motor will use less power than a 500 watt, but it produces less power too. and the reason for having a motor is to produce power. In this case, it doesn't matter that the smaller motor is geared. its a less powerfull motor, it uses less power.

The truth about Geared motors is they are more efficent at accelerating, but less efficent at crusing speed. Most geared hub motors fall in the 75% efficent peak range. A direct drive can be over 80% peak. If you compare a 500W BMC/MAC to something like a 500W 9 Continents on http://www.ebikes.ca/simulator/ you'll see that the peak efficancy in that case is 8% less for the geared motor. But you will also see the geared motor's efficancy curve rises faster, meaning it gets nearer its peak efficancy at lower speeds.

Most of a Geared motor's efficancy loss comes from drivetrain losses. There is no such thing as a frictionless drivetrain.

 

[SamTexas]

Dec 28, 2011
I have a love/hate relationship with my 9c 2807 DD hubs. I like them because they have no mechanical parts and because they can take more power. I hate them because they are so inefficient. I have done many tests comparing the 9c 2807 with the cell_man MXUS 350w geared kit. Running on the same test loop, on the same bicycle, maintaining the same riding style and average speed, the 9c 2807 always consumes AT LEAST 20% more energy.

I would like to know exactly (or as exactly as possible) the source of that inefficiency and its magnitude at various speeds. It appears to me that the higher the speed, the less efficient it gets. What kind of test, what kind of instruments do I need to get solid, repeatable scientific numbers?

Thanks.

they are compairing apples to oranges. I happen to have both kits from them. I actualy get a little further on my direct drive kit.

But they are comparing a small motor to a large motor, of two diffrent technologies. Its like saying a 40 watt lightbulb will save more power than a 75 watt bulb. True, but that ignores the reason for having a bulb. Light. and the 75 watt will give more light. a 350 watt motor will use less power than a 500 watt, but it produces less power too. and the reason for having a motor is to produce power. In this case, it doesn't matter that the smaller motor is geared. its a less powerfull motor, it uses less power.

The bold and underlined text in my above post indicated that I was comparing apples to apples.

 

[neptronix]

The efficiency difference has to do with speed.
In fact, it is not an efficiency difference; lower speeds just take less energy

Geared motors are usually drastically more efficient in hilly situations, but not as good on the flats, 5-10% less efficient.

That being said, i'm a big geared motor fanboy since i live in a hilly area, and my bike gets stored upstairs, so the low weight is a huge plus. I also like the stealth effect - very few people notice that my bike is electric.

 

[Drunkskunk]

A 350W geared motor compared to a 500 watt direct drive is not an apples to apples comparison.

And My MXUS will get me around 8 miles. Using the same battery, my 2807 gets around 10.

 

[John in CR]

Geared hubbies have lower peak efficiency due to the losses in the gears. Due to the big step down in gearing the geared system is better able to operate in the meat of it's efficiency curve. Unfortunately the vast majority of DD set ups are too steeply geared for their load and voltage or have incorrect current limits, so they rarely if ever get up to their prime efficiency range of operation. Type of riding can have a huge impact on overall efficiency and range.

If you're really worried about efficiency, then put a DD in a mid-drive and give it variable gearing. LIghtcycle reported a 50% increase in overall efficiency, and that was with a NuVinci sucking down at least 10% in losses, that it more than made up for by providing variable gearing.

For a range competition I'd go with a DD hubbie any time, and put it in a smaller wheel run at the voltage needed for the speed I wanted. I speed wind hubbie in a small with the amps severely restricted has the best power/efficiency curve of all the single speed systems we run. Too bad no one runs them.

 

[dogman dan]

John has it correct as usual.

Sam, I just have to think that your test failed to completely eliminate all the ride style variables. It's not so easy to ride 20+ miles the same way twice. Or the weather could be different. Enough stops would certainly skew the results towards the gearmotor. My tests were done on routes with few stop signs.

How many times did you repeat that test? I never saw any big difference in my results unless something else changed, ride speed, stops, weather, or battery. I did not do just one test. One test was done in carefully controlled conditions, but then daily commuting for over a year confirmed the results. In similar weather, 20 mph travel used very similar wh.

But for sure, every single time I uncorked the full power and speed of the DD motor, it resulted in worse wh/mile numbers. My personal best distance on a 36v 20 ah pingbattery with the small gearmotor was 35 miles. Personal best with the same battery and dd motor is 39 miles. But surely conditions and ride speed were different enough on each of those particular rides, so that data is pretty meaningless. But on many many many other rides, both types of motors typically went exactly 30 miles when trying to ride 20 mph.

 

[SamTexas]

Sam, I just have to think that your test failed to completely eliminate all the ride style variables. It's not so easy to ride 20+ miles the same way twice. Or the weather could be different. Enough stops would certainly skew the results towards the gearmotor. My tests were done on routes with few stop signs.

How many times did you repeat that test? I never saw any big difference in my results unless something else changed, ride speed, stops, weather, or battery. I did not do just one test. One test was done in carefully controlled conditions, but then daily commuting for over a year confirmed the results. In similar weather, 20 mph travel used very similar wh.

The comparison tests were done 2 times (2 runs for each motor, 4 runs total). The route is the same: 18 mile loop in the same direction, I was the rider. The average moving speeds are the same (+/- 0.2mph). Top speeds are almost the same (+/- 2mph).

 

[spinningmagnets]

On flat ground, I don't think there's going to be a big difference in efficiency (from what I've read). Most geared hubs have a ratio of around 5:1, so the rotor is spinning about 5 times faster than a DD at a given speed.

If you take a low-turn (speed) DD hub, and try to run it up a steep hill, it will turn a lot of your battery watts into heat (worst case scenario). But...if you have a high-turn count DD hub and you run a mild hill, you will find out if the motor can stay at its top speed without bogging down (staying in its efficiency range), and also see if it can shed the heat being made as fast as it is being generated (on the hills of the particular steepness/length that you have). If yes, the efficiency shouldn't be much worse than a geared hub on the same ride.

There is a limit to the off-the-shelf options available, so...it depends on what your commute is. A single steep hill, many small hills, flat ground?...20-MPH tops, or 30-MPH+?

If you have a lot of starts and stops, and a top-speed of 26-MPH or less, and mild hills, I'd recommend a geared-hub (8T 500W from cell_man @48V).

If you have few stop-n-go's, spend most of your time from 15-35 MPH, no steep hills, I think you may be better served by a DD. I read a lot, but I've never owned a hub, so take anything I write with caution.

 

[Russell]

I did a test between a little 250W Bafang geared motor and a 9C direct drive motor a couple of years ago. I traveled the same route at the same average speed and ya know what...the energy consumed was virtually IDENTICAL.

When cruising at full throttle at the same speed a DD motor should return a slightly higher effciency however if the motor is being used as assist then a lighter freewheeling geared motor, especially a slower wound version, is the best choice to extend range. The motor isn't more efficient however the cyclist is likely to pedal more often without the motor yielding better range.

-R

 

[Sunder]

So which type would be most efficient for pedal assist rather than all motor? I know that free wheeling geared hubs are best when you want to pedal, due to lack of resistance, but what if your mode of motion is 70% human with the motor just topping off your speed?

 

[ddk]

So which type would be most efficient for pedal assist rather than all motor? I know that free wheeling geared hubs are best when you want to pedal, due to lack of resistance, but what if your mode of motion is 70% human with the motor just topping off your speed?

are the best for that use, if you're riding somewhat sane bicycle speeds (5-15MPH)

 

[dogman dan]

I'd have to be guessing on that one. I can't say I ever did much testing where I was using less than 50w of motor power. It's extremey difficult to hold a throttle on a 20 amp controller much lower than 100w. So for me to put out 70% of the power, I'd have to pedal somthing like 200w of the 300w total. That's too much pedaling for me to do continous.

I have done a few long rides where the motor was held at about 100-150w, and I pedaled up another 75-100w. Those were my personal best marks. But I never repeated that enough times to have the data. One test with additional variables is junk science data. For sure, you go a long ways if you pedal up 30-50% of the wattage. At this point, your range is so long that which motor is used could well be a moot point. Either one goes so far it doesn't matter which is 5-10% better. I do NOT believe in gearmotors being 20% better unless the route includes many many stops.

But in my mind, I would not expect the gearmotor to have a big advantage on open roads untill you are riding with the motor completely off for some of the time. Pulse and glide riding, where you travel perhaps 10-15 mph, using the motor only for the uphills, is definitely using less power over the whole ride.

This is often called more efficient. But splitting hairs, I call it riding with no motor. Subtly different from more efficient, it is in fact, just more pedaling. If you are going to ride pulse and glide, basicly having the motor off 70% of the time, then of course the freewheeling motor is the only way to go. But splitting the hair, it's not a more effcient motor, but a motor that pedals better.

 

[Sunder]

Thanks Dogman

Re: being difficult to hold down power, I can put out about 160-180w for about 20 mins so maybe less of a problem, but its also part of the reason why I'm looking at a lower powered motor (Cute Q100) From the graphs I've seen, motors are very inefficient below about 50% throttle, but I also don't know what their efficiency is like if you over drive them substantially.

I think my problem is that I want the bike to do everything well and be cheap in the bargain too. I don't think that's gonna happen.

 

[cal3thousand]

Thanks Dogman

Re: being difficult to hold down power, I can put out about 160-180w for about 20 mins so maybe less of a problem, but its also part of the reason why I'm looking at a lower powered motor (Cute Q100) From the graphs I've seen, motors are very inefficient below about 50% throttle, but I also don't know what their efficiency is like if you over drive them substantially.

I think my problem is that I want the bike to do everything well and be cheap in the bargain too. I don't think that's gonna happen.

I think we all want that. Same with cars sometimes. I want to get the best gas mileage, have power when I want it, utility, be affordable, sustainable, etc.

 

[amoras]

Direct drive would have 15 to 25% less range under the same condition? that's huge.

Why don't you use my ERS (Ebike Route Simulator) and check it out
http://endless-sphere.com/forums/viewtopic.php?f=3&t=37790&start=15

You will need a torque curve for those motors or instead use a similar one from the ERS.
Start comparing n.º 20 with n.º 23.

The below statement is for ERS spreasheet, using similar setups with the almost the same unload speed and default route (n.º 20 and n.º 23 - low winding setup):
Rules of Thumb
1. The DD wins up to 3% of grade and looses above that.
2. The Wh/km will be lower for DD then geared. How much? Not much, around 10%.
3. The Geared drive will always be faster.

Apply your route to ERS and see if the above statements are true for you.

 

[Alan B]

On my very hilly commute the Geared motor on a Recumbent is notably more efficient than a DD on a mountain bike.

Note that more than one variable was changed, so attributing the proportion of effect is not differentiated.

When using little motor power the freewheel of the geared motor saves a lot of power over a long ride. They glide much better than DD. Justin's measurements showed the difference was similar to the difference between good and poor tires. It seems like more than that when you are pedaling. If you are going to be pedaling a lot without motor power the geared motor with a freewheel wins.

When gliding downhill with geared motor on recumbent I find that it glides more efficiently and overtakes road bikes. The mountain bike with DD glides much less efficiently than a road bike.

Low turn count "speed winds" have slightly lower efficiency than high turn count motors due to the higher proportion of ineffective copper outside the teeth.

Matching motor speed to application generally produces best efficiency.

The DD is quieter than the geared motor, but neither are silent nor are they all that noisy.

The heavier DD has more impact on the weight of the bicycle, so the geared retains a lighter more "bicycle like" feel, and the freewheel contributes to this. The DD feels heavier, harder to pedal and slightly less "bicycle like".

I like them both, they are different tools with similar but not equal application utility.

 

[o00scorpion00o]

I installed my old conhismotor motor today and used the same battery and controller as I used with the pie.

The conhismotor 48V 1000W is supposed to be the same or close to the 9C,

So after testing the Pie, Conhismotor, and mac, the mac beats the shit out of all of them! The mac 8T @3500w has more torque and speed than the conhismotor at 4800w on 16S LiPo.

The Pie @ 4800 watts feels as if it has as much torque as the mac at 3500 watts, but the mac goes to 41 mph max, the pie 30 sensorless and 35mph with halls!

The pie is more efficient at 30 mph than the conhismotor!

The mac climbs hills faster and is more efficient at 30 mph than the conhismotor and hard to tell the difference with the mac.

 

[cwah]

So in term of efficiency under acceleration we have the Mac > Pie > Conhismotor? (what about the 9C? )

Because it's geared, under acceleration the mac is expected to perform better.


However, are the direct drive also more efficient on constant speed as expected?

For example at 20mph constant or 30 mph constant I suppose the direct drive should theorically use less watt than the Mac.

 

[o00scorpion00o]

So in term of efficiency under acceleration we have the Mac > Pie > Conhismotor? (what about the 9C? )

Because it's geared, under acceleration the mac is expected to perform better.


However, are the direct drive also more efficient on constant speed as expected?

For example at 20mph constant or 30 mph constant I suppose the direct drive should theorically use less watt than the Mac.

The conhismotor is almost identical to the 9C, but I'm not 100% certain. There are people who believe it is. They are all much the same.

The Pie is pretty efficient at slower speed, the mac is also very efficient. @20 mph I managed to get 25 miles from just 8.5ah no peddling, with a few hills with the mac. I have my tyres pumped to 90 psi on all motor tests and use skinny 1.6 Continental sport contacts. I like maximum pedal efficiency and those tyres make a huge difference compared to normal mountain bike tyres, so It can only make a difference to motor efficiency!

I did notice the conhismotor the least efficient out of the 3 motors. At 30 mph the conhismotor was consuming about 300-400 watts more than the pie, the pie consumes about 800-1000 watts on level ground, slight breeze.

I don't think there would be much of a difference between the pie and the mac at 30 mph, it's hard to tell the difference.

But that's where the similarities end, the 8T mac produces the around the same torque as the pie with about 1000-1300 less watts! There is absolutely no doubt about that, it's astonishing really when all 3 motors are together and you see the difference!

The mac is also the quietest by far of all 3 motors, the direct drives make a very loud growl and people certainly notice you under acceleration. The pie makes a higher pitch at certain speeds but wind mostly drowns that out, the direct drives make a noise at higher speeds but fairly quiet, the conhismotor resonates at certain speeds.

I should also add the mac needs the least voltage of all three motors for 30 mph, the 8T mac being the only motor out of the 3 that can reach 41 mph from 16S LiPo, Although the conhismotor would get pretty close if using halls and 120% in the controller like with the mac. The pie needs the most voltage!

Here is the Conhismotor before installation and before I replaced the phase wires and tyre.

The Pie V Mac stator