Efficiency test Geared vs Direct Drive Hub motors

Did a test ride, 6.3 miles, same route and speed, same battery, one bike with geared front hub, the other with direct drive front hub, same rider.

The geared hub was a front Bafang G311, a small 11:1 reduction gear hub, on a 24" wheel Townie, with a 32 x 14 top gear and a 36v, 10ah battery charged to 42.0v. Rider weighed 210 lb total and pedaled on the flats and climbs. Average speed was 10.6 mph.

The direct drive hub was a front Aotema, on a 26" Jamis Hudson, much like a Townie, with 42 x 14 top gear but downshifted a couple cogs to match the gearing of the Townie. Same battery and rider were used, average speed was also 10.6 mph.

Result: voltage final was 40.1v for the gear hub Townie and 39.8v for the direct drive Hudson. That's not a big difference, charge time was less for the Townie run than for the Hudson, as would be expected. 0.92 hrs vs 1.03 hrs, both at 2 amps and titrated to a full 42.0v charge. While not a big difference it would give the Townie about an extra mile of range, using battery voltage from 41v to 35v. I expected a bigger difference, the actual advantage was probably mostly due to tires rather than hub motor type.

Tires were Kenda K-838 24" x 1.95 on the Townie and Performance Gotham 26" x 1.75 on the front of the Hudson and no name tire on the rear.

Coast down testing of my bikes, overtaking my wife's bike, pedal only, no hub motors, revealed that Kenda 838 26" x 1.95 on my FS bike versus Gotham 26" x 1.75 on another of my bikes, showed the 838's have significantly less rolling resistance than the Gothams.

Well, I was wondering about geared hub vs direct drive and got a surprising answer.

Both will have a slight advantage over each other in various situations, except for reliability and top speed that are definitely DD hub domain. They just serve different purpose but the DD hub can be used in Geared hub territory, while the opposite is bound to fail.

Very similar to what I found when I first started trying different motors. At same speed, same route, same weather, and few stops, the efficiency difference can be made up with a few yards of extra pedaling on the DD. Makes sense, both are very efficient, brushless motors.

But try another test, where you have a stop sign every block, and start without pedaling. After all, you aren't testing pedaling, but the motor. You will find the geared motor shines over the aotema then. That first 5 feet the geared motor has a huge advantage.

Solution,, pedal the first 5 feet on the aotema, then start the motor when going 10 mph. There are easy ways to improve efficiency, if you have the wrong type for your rides.

Re geared motors.. I have been pounding pretty hard on a larger geared motor that E bike kit sent me to ride till it broke. I could overload and kill it in 30 min. But in normal use, riding dirt single track, I just have not been able to kill it. That motor is going on 4 or 5 years now. They just don't break the gears like they did in 2008 anymore. But they can be killed deliberately, with overload, or over volt.

Efficiency of a BLDC (or any electric motor) drops at low speeds.

At 0 RPM no matter how much power you throw at it above 0 Watts, the efficiency is..........Zero.

Bike wheels don't rotate all that fast; indeed a 20" rolls faster than a 26" but we're dealing with what, 150-300 RPM typically?

Efficiency rises from 0 as poles are switched.

Consider a larger DD motor with 48 pole switches/rev vs a Geared motor with 160 pole switches/rev, which do you think is most efficient below about 50% max speed? On the average its not until about 60% max RPM that a DD motor equals a geared motor efficiency and its not all that much better at higher speeds. Granted there are a number of variables that alter those numbers like voltage and actual speeds but for the most part that applies to about 90% of the configurations out there.

A lot depends on how one rides:

• Flat paths with average speeds over 18-20 MPH is where you need to be for a DD to equal a geared motor on efficiency

• Hills where speeds drop to 15 MPH or below a geared hub will be far more efficient, and dropping to 10 MPH a geared hub like a MAC 10T will nearly double the efficiency of any DD hub on the market

I did quite a bit of testing with several motors a few years ago for a Trike project and here is an excerpt from what I discovered (and published).

Efficiency of the Bafang BBS02 Mid Drive is head and shoulders above the Geared or Direct Drive motors tested. On the same 12 mile route with over 5 rides on each motor and maintaining the same average speed (14 MPH) the total watt-hours consumed by motor:

BBS02 - 68 Watt-Hours
MAC 10T - 91 Watt-Hours
Magic Pie 4 - 123 Watt-Hours
Smart Pie 4 - 130 Watt-Hours
This course includes 336 feet of altitude change, several standing starts up hills and some roller coasters with 610 feet total climbing. I do pedal typically producing about 100 watts of power as measured by the Cycle Analyst. There was little variation of Human Power from the Hubmotors, however the Thun Torque Sensing BB cannot be used with the BBS02 so there is an assumption the Human Power is about the same. If its not, than its certain the BBS02 makes a rider more powerful.....


There were 3 other DD motors I also tested all having the same 12 FET external Trap wave Infineon driver that was used with the MAC and they weren't far off the Magic motors although a little lower in efficiency. Also not included was a Go-Swiss which was slightly better than the Magic Pie, mostly due to the higher pole counts.

At the end of testing over about 4 months I settled on the MAC 10T with a Phaserunner coupled to a Thun BB. I expected the Sinewave to burn more juice, it actually upped the efficiency due to the FOC (FOC was also in the Magic Pie and Go Swiss motors).

Unless you have long steep hils like I do, you won't really gain that much in regen. As speed drops so does both voltage and current.

As they say YMMV.

Link to the MAC 10T test:
http://www.triketech.com/Drivetrain/PowerAssist/HPV-MAC-V2.html

FWIW, I've spent a few years designing motors, including the AC Asynchronous drives used on the Terex MT6300 Mining truck (arguably still the most powerful hubmotor at 4.2 Megawatts) as well as precision positioning gimbal motors that can lock onto a target with 4096 "steps" between the 12 poles they had along with the drive system based on a TI 8312DRV and C2000 PIC.

More mass with the direct drive hubs makes it easier to shed heat. Another + for dd's.
No moving parts, another + unlike the gears.
Less parts for dd.
DD's are cheaper.
DD's are heavier.
DD's can take more abuse, and more power.

Any kind of load that slows a motor a lot can push it to inefficiency. That is why the stop and go ride can be up to 20% less efficient when comparing some DD motors than a geared.

Solution, pedal first. Or,, more power so the motor never slogs. Its the temporary overload that gets so inefficient. Unlike cars, a jack rabbit start on electric could be more efficient, compared to a temporarily overloaded, less powerful motor.

It's simple, 0 rpm for the shortest possible time. You eliminate 0 rpm load by pedal first, or make the low rpm period shorter with more HP.

Worst condition of course, is overload a small motor up a mountain. It will run inefficient enough to melt. In general, if your hub motor can't run you up a hill at 15 mph, with moderate pedaling help, then you or your cargo weigh too much for the motor and the grade.

Both rides included 10 starts from full stop and 4 low speed roll throughs. Two roll throughs were over short but high speed bumps.

With both motors I started using very little throttle, as is my normal practice, in an attempt to limit amp draw at zero and low motor rpm. The tiny geared motor does have stronger acceleration from zero to 8 mph. Both rides were at much lower speed than I normally ride the direct drive. In climb testing, the direct drive motor temperature is only a bit above ambient temperature, but I'm only using a little electric assist, most power comes from pedaling.

The big difference in low speed efficiency is a strong argument for a geared motor, for uses like Moab Slick Rock Trail or the Practice Loop. Yeah, I'm stuck on the Moab bike thing. Moab is very hot over the slick rock in the summer. Heat would be a problem for the components during June through September. The Practice Loop is about 2.5 miles long, the Slick Rock Trail is more like 13 miles.

Present thinking is use a rear geared hub motor on my FS bike and pedal with almost as much effort as pedal only riding. A second option would be to use a front suspension hard tail, with suspension seatpost, and a rear geared hub motor. Plan is to use minimal electric assist while pedaling pretty hard. Ride speed would be low and walking the steep parts would be an embarrassing option. Coasting downhill should be excellent, again at low speed.

I was expecting to read that you were going to go by way of 2wd with two hubbies, which is less efficient.

I was also expecting maybe to read that you were going to go geared hub rear + dd hub front, and switch mid ride, or use both at the same time. More exotic, more weight, more complicated.

I've seen a rider here in San Antonio who was using both front and rear hub motors. Got to talk to him once, he knew a lot more about it than I do. When I've seen him he isn't riding fast, just riding at around 12 mph.

You will always have more performance by weight, with one big motor than 2 small ones. The reasons to build dual motors are not performance, power or speed. I’d say, most of those building with dual motors are doing it for the wrong reasons, and valid reasons are very few.

Agree. One big motor, keep the front wheel lighter.

For moab, a larger rear geared motor would be ok, but a lot better would be a mid drive.

I ride my desert trails fine with a 500w geared motor, but really, don't go riding it on hard trails when its all that hot. Dawn patrol, or just go to 9000 feet where its cool in June. Retired, so I get out of the hot desert in summer a lot now. 1000w to the motor, 52v.

A big ass motor would shine in moab, but I'm not so sure they allow it on the trails. One of the reasons I took my big ass motor off the dirt bike, and went to the 500w geared, was so I could legally ride it on the local national monument trails.

For the rest, I bought a small honda, for riding with road or trail damaging power. I built miles of trail very close to my house for my high power e bike, and it serves well for the small motorcycle.

If you have a lot of time, stop by the organ mountains national park, for a warm up on the way to moab. Lots of good trails here, just not as spectacular as Moab. It should be on your route, about half way.

Dogman Dan, what do you think, rear MAC 10T for the Moab Slick Rock and Practice Loops? Or the 12T?

Bike options would be 26", either my FS or a hard tail with suspension forks and suspension seatpost. The FS is in the garage waiting, but I'd have to buy a hardtail with suspension fork and get a good suspension seatpost. So, the FS would be far cheaper.

This would probably be a one time thing, but maybe not. The Moab bike would also be used on the trails around here, very mild stuff. My riding is closer to crawling than mountain biking. Thinking of keeping the Hudson as the grocery getter bike.

Retired here too, 73 going on 14. Well, maybe not the 14 part.

MC is a Yamaha XT 225, it's great for the street but too big and heavy on the trails for me. For one thing it's too tall and I seem to be getting shorter. The old Honda Trail 90 or 110 would be a lot easier to pick up than the XT. Short legs is sort of a problem with the FS bike too. Love the suspension on the XT though.

Best Wishes and thanks for the help,

Mike S

I'd be interested into reading the same kind of stuff with 20x2.4" wheels or 20x3"... It's kind of absurd to used large wheels with hub motors, so why not test/compare them where they truly shine

qwerkus said:

I'd be interested into reading the same kind of stuff with 20x2.4" wheels or 20x3"... It's kind of absurd to used large wheels with hub motors, so why not test/compare them where they truly shine

Click to expand...

You mean large diameter rims laced to hub motors. Heats everything up, less torque, less get up and go speed, higher top speed is about the only pro. Going down one size of wheel, like from a 26" bicycle to a 19" motorcycle which equals 24" bicycle will net higher torque, less top speed, more get up and go power, less heating up of the motor/controller which means you can push more power through the motor.

Question is, how much more torque is there from 26" bike to 19"mc=24"bike and how much more get up and go power.
Going to a wider tire nets more traction, but more friction too but the loss of Ah is neglible.

Well, large wheels are a must in some situations. I mean, try riding fast in mountain trails with small wheels. Your ride is turning into a rodeo at the first series of small bumps. That is one reason why there are slower motors, bigger motors too.

Small wheel and fast motor, is the fastest, most efficient solution, only as long as the small wheel does let you perform that speed on the terrain. Then, motorcycle wheels that are only 2 inch smaller, are not giving much advantage because of the weight. They will give better efficiency only if you smooth out the acceleration, because their inertia is much higher. And, this higher inertia will make a loss of braking performance, slower steering, slight under turning, complication in suspension tuning...

Only advantages I see in using
motorcycle rims - can ride on the rim longer to get home, oh yes I have done that a few times.
motorcycle tires - more durable, less flats

Ideally want that magic motorcycle tire on a bicycle rim, if rim fits bead of tire.

This gives details on speed and torque and such of diff size dia wheels in hubbies. But as a seat of your pants feel, be nice to know myself.
https://www.ebikes.ca/tools/simulator.html

Yep. Most who are using motorcycle wheels on ebikes, are doing it because they tired of repeated flats. Some regions are really hard on tires, because of thorns or transport debris. I can’t recall Canmore being such a problem, but it’s been a long time since I’ve been there.

markz said:

Ideally want that magic motorcycle tire on a bicycle rim, if rim fits bead of tire.

Click to expand...

https://www.daekekspert.dk/en/Products/TyreByKey/?ProductKey=15079189

22” moto rim = 26” MTB rim

Balmorhea said:

markz said:

Ideally want that magic motorcycle tire on a bicycle rim, if rim fits bead of tire.

Click to expand...

https://www.daekekspert.dk/en/Products/TyreByKey/?ProductKey=15079189

22” moto rim = 26” MTB rim

Click to expand...

Maybe for puncture proof, not for speed. I mean, they are rated 100 kmh. I ride faster than that on bicycle slicks for years and never blown one. We also beat that speed on DH racing tires. Not saying those moped tires are not good, just that I would like some info on gum hardness, weight, and how hard they are to mount on a 26’’ bicycle rim.

The shortcoming of any motorcycle tire is efficiency, as compared to a bicycle tire.

That moped tire may only be rated for 100kph (62mph), but I see plenty of bicycle tires that boast 50kph (31mph) ratings, that I know are good for at least twice as much speed if you stay within their weight ratings. Those 50kph rated e-bike tires are probably sturdier than the non speed rated bicycle tires you use at over 100kph.

I don’t want to squander my muscle power or battery power on a heavy, poor riding motor vehicle tire. But if I did, I would not fool myself into thinking that a low speed rated moto tire was less speed capable than a bicycle tire.

Balmorhea said:

The shortcoming of any motorcycle tire is efficiency, as compared to a bicycle tire.

Click to expand...

I still get 25Wh/mile-30Wh/mile either way.....

on my big DD hub pedaling PAS limited 750w, with much stronger heavier 75mph MC rim/tire... (9.9lbs) incld. axle
vs ...
the 750w BBS02 @ 22A or so.... PAS on 26" bicycle tire MTB geared bike. (5lbs incld. axle...)

The bead, rim strength, the speed rating on the face of the tire along with the manufacture date, cost of replacement, durability, spoke strength, all outweigh the trade off of 4.9lbs....

You ever get those weeks where you have one flat after another? I hate that. Have not had that with a proper beaded rim. I beat on it. I like a strong tire, and tires make the ride in alot of things. Bicycle tires get beat up to fast and then they floppy weak and loose.. or they may be really old when you buy them.... Old stock.

Did you go with a HD or UHD tube?

Lots of flat tires, are apart of my riding experience as of late.

DogDipstick said:

You ever get those weeks where you have one flat after another? I hate that. Have not had that with a proper beaded rim. I beat on it. I like a strong tire, and tires make the ride in alot of things. Bicycle tires get beat up to fast and then they floppy weak and loose.. or they may be really old when you buy them.... Old stock.

Click to expand...

New DH racing rims and tires are tubeless, beadless, and we ride them with Cush Core inserts and slime. Better flat protection than most motorcycle tires. And they are softer gum than most motorcycle tires too, giving the extra grip to ride fast on a light weight bike. They have a short life of course, but that is the price to pay for a better ride.

The aotema motor is pretty inefficient as far as DDs go and the bafang is a newer higher efficiency design. Your result isn't a big surprise

ebikes.ca motor simulator shows us that geared motors generally lag well behind modern DDs in efficiency in a majority of situations..
https://www.ebikes.ca/tools/simulator.html

Most efficient geared motor i know of is the newer MACs with 0.27mm laminations.. they may peak out at about 84% efficiency.
Most efficient DD might still be the 35mm wide "1500w" leaf motor.. peaking at 90.5% efficiency but with a very broad efficiency curve.

markz said:

Did you go with a HD or UHD tube?

Lots of flat tires, are apart of my riding experience as of late.

Click to expand...

The lightes smallest one. The HD. Not the heavier one. Working fine.