Hub or Mid Drive? Any views?

[danielrlee]

Another disadvantage of a mid-drive over a hub motor, is the interruption in torque delivery each time you reach the motor's BEMF and have to change gear as you work your way through the available speed range. Hub motors don't suffer the same issue and deliver continuous torque through their entire speed range.

A hub motor's torque does progressively decrease to zero in the upper half of its speed range due to BEMF. However, this still compares well to a mid-drive, where torque in the upper gears is naturally lower due to the effects of 'gearing up'.

 

[Woosh]

Another disadvantage of a mid-drive over a hub motor, is the interruption in torque delivery each time you reach the motor's BEMF and have to change gear as you work your way through the available speed range.

depending on the motor but typically, you have to pedal at well over 100 - 120rpm to start losing power on a crank drive.

 

[anotherkiwi]

Within the legal limit (25 km/h) of assistance about 100W is needed to power a racing bike on the flat.



This is a graph for a 5% gradient, the kind of hill most people have to negotiate on the road, and that requires about 420W. I have managed to sustain 25 km/h on a 9% hill on a hub motor bike - 540W on the LCD + the Watts provided by me.



This shows the importance of being in the correct gear - even on a hub motor bike - to make the best use of the power available.

 

[anotherkiwi]

depending on the motor but typically, you have to pedal at well over 100 - 120rpm to start losing power on a crank drive.

From personal experience about 100 rpm on a 36v GSM mid motor. I don't have back EMF at 42v and above.

 

[danielrlee]

depending on the motor but typically, you have to pedal at well over 100 - 120rpm to start losing power on a crank drive.

Yes, but that's missing the point somewhat. With a mid-drive motor, torque delivery has to stop regardless whenever you change gear.

 

[Woosh]

most CD bikes now have gear shift sensing, the effect is tiny.
I feel that the problem with CD bikes is age lowers cadence.
Outside the optimal range (70-100 rpm), you negate the benefit of a CD motor.

 

[danielrlee]

This shows the importance of being in the correct gear - even on a hub motor bike - to make the best use of the power available.

The difference is that with a hub motor, an "incorrect gear" will only affect the human input power. With a mid-drive motor, both human power and motor power is affected.

 

[anotherkiwi]

The difference is that with a hub motor, an "incorrect gear" will only affect the human input power. With a mid-drive motor, both human power and motor power is affected.

My "only human input" is between 160-200W so I tend to notice wrong gearing more than someone putting in less effort I guess. Even at 100W human input being in the correct gear with a hub motor will extend battery range by a significant amount.

I have stalled my mid motor once in bottom gear. That told me I needed to redo my gearing because unfortunately hills like that are not uncommon where I live (the Basques say "mountains" but they tend to exaggerate...).

 

[sjpt]

A couple of questions I've asked before but not seen (or understood) answers?

Is there any reason a crank motor is inherently best at 70-100rpm? Could a motor not be made with different windings, or different in some other way, that is optimized for cadence of (say) 50rpm.

Is there any reason not to have a hybrid torque/cadence sensing system? It would need careful design mediating the two; but at a first guess just use the maximum that the two systems suggest at any one time. The I'd get the advantage of 'natural' torque sensing and good immediate power at junctions, and the ability to lazily ghost pedal and have the motor do all the work?

~~~
A comment for those into the figures: most of you probably already know. There is a really useful page at https://www.ebikes.ca/tools/simulator.html that shows various features of different motors under different conditions; unfortunately I'm not sure if it has been updated with some more recent motors.

 

[anotherkiwi]

1. the manufacturer could modify the internal gearing, this was done recently with the Bafang G310 hub which has high first stage rpm and 11:1 reduction final. There is no reason not to apply the same to a mid motor

2. https://www.bafang-e.com/en/oem-area/components/component/motor/mm-g30200c/
I think the Bosch motor may also use torque and cadence sensors, it has been a very long time since I read the tech specs for those motors.

 

[Woosh]

torque based CD motors use their cadence sensor too, firstly to stop false triggering at traffic lights. You have to rotate the cranks a bit to start the motor. Secondly to give a bit more push at low cadence.

Is there any reason a crank motor is inherently best at 70-100rpm? Could a motor not be made with different windings, or different in some other way, that is optimized for cadence of (say) 50rpm.

crank motor and hub motor differ mainly through their gearbox designs, substitute 70-100 rpm at the crank by 200-300 at the wheels, you get to the same conclusion.

 

[vfr400]

depending on the motor but typically, you have to pedal at well over 100 - 120rpm to start losing power on a crank drive.

If you have a crank motor that still gives power at 120 rpm. It'll be very inefficient at 60 rpm, which is the sort of cadence many of our forum members have. You'd only want 120 rpm for for sports riders.

 

[Nealh]

I can only wish to have a high cadence, at best 74rpm is about my best but not sustainable for miles.

 

[vfr400]

A couple of questions I've asked before but not seen (or understood) answers?

Is there any reason a crank motor is inherently best at 70-100rpm? Could a motor not be made with different windings, or different in some other way, that is optimized for cadence of (say) 50rpm.

Is there any reason not to have a hybrid torque/cadence sensing system? It would need careful design mediating the two; but at a first guess just use the maximum that the two systems suggest at any one time. The I'd get the advantage of 'natural' torque sensing and good immediate power at junctions, and the ability to lazily ghost pedal and have the motor do all the work?

~~~
A comment for those into the figures: most of you probably already know. There is a really useful page at https://www.ebikes.ca/tools/simulator.html that shows various features of different motors under different conditions; unfortunately I'm not sure if it has been updated with some more recent motors.

There is a difference between different crank motors on the market. Some are very easy to rev out, which is really annoying if you have a high cadence or established pedal routine. The problem is that if you increase its max rpm for people with a high cadence, you lose efficiency for those with low cadence.

 

[anotherkiwi]

I can sustain 105 rpm for about 30 km, I haven't tried longer distances on the flat in quite a while. Any cadence below 90 I feel there is something wrong - either with me or with the gear I am in. Climbing very steep gradients I get down to about 60 rpm.

 

[RobF]

The Bosch system uses cadence, torque, and speed sensors.

One of it's strengths is 'set and forget' - put it on tour setting and ride it like an ordinary bike.

I take the smooth power delivery for granted, but it does make some other crank drives feel clunky.

Back to David's new bikes, which crank drive motor is in the crank bike?

 

[sjpt]

Clearly active cycling enthusiasts have much higher cadence. However, a significant part of the ebike market is older people with lower cadence; many never rode with high cadence, some (me) used to but don't feel so like it now. A single crank motor can't be optimized for both as vfr400 says, but I would have thought there was a large enough proportion of the market to have motors optimized for 50-60.

That said, looking that the Bosch range assistant (https://www.bosch-ebike.com/en/service/range-assistant/) there does not seem to be much efficiency difference over quite a wide variation. With some 'typical' (for me) parameters I see a range of 32 miles at 50rpm against a max of 35 at 100rpm.

I see from the answers that many torque based systems are indeed cadence as well. Should be easy to change the algorithm to allow powered ghost pedalling ... I'd certainly like it and I would have thought others would too. It could still be 'set and forget'. Each of the settings would behave as now with significant user torque, but give a minimum power (as long as there was some pedalling); say 0% (eco), 20% (tour), 50% (speed), 80% (turbo).

 

[vfr400]

Clearly active cycling enthusiasts have much higher cadence. However, a significant part of the ebike market is older people with lower cadence; many never rode with high cadence, some (me) used to but don't feel so like it now. A single crank motor can't be optimized for both as vfr400 says, but I would have thought there was a large enough proportion of the market to have motors optimized for 50-60.

That said, looking that the Bosch range assistant (https://www.bosch-ebike.com/en/service/range-assistant/) there does not seem to be much efficiency difference over quite a wide variation. With some 'typical' (for me) parameters I see a range of 32 miles at 50rpm against a max of 35 at 100rpm.

I see from the answers that many torque based systems are indeed cadence as well. Should be easy to change the algorithm to allow powered ghost pedalling ... I'd certainly like it and I would have thought others would too. It could still be 'set and forget'. Each of the settings would behave as now with significant user torque, but give a minimum power (as long as there was some pedalling); say 0% (eco), 20% (tour), 50% (speed), 80% (turbo).

How does that Bosch simulator know how hard you're going to be pedalling? It has no idea of who I am, how much I practice and how fit I am. Sometimes I pedal hard for the whole journey, and sometimes I don't pedal hard at all.

 

[Woosh]

How does that Bosch simulator know how hard you're going to be pedalling? It has no idea of who I am, how much I practice and how fit I am. Sometimes I pedal hard for the whole journey, and sometimes I don't pedal hard at all.

you can always find exceptions but that does not negate the value of a carefully designed piece of software to predict desired power given cadence and speed.
It's just like speech recognition, a dictionary helps but it's mainly pattern recognition and statistics.

 

[vfr400]

you can always find exceptions but that does not negate the value of a carefully designed piece of software to predict desired power given cadence and speed.
It's just like speech recognition, a dictionary helps but it's mainly pattern recognition and statistics.

But I can't find where you input your own power. That's the biggest single factor that affects range. Without it in the algorithm, any prediction could be wildly out. Some people pedal consistently at 200w, others struggle for 50W.

Lets say it calculates a range of 64km (40 miles) and has the 400wh battery. That's 10wh/m or an average of 120w from the motor at an average speed of 12 mph. We have no idea what power they've guessed at for the rider, but lets say 100w, so that range calculation would be based on an average of 220w, but the difference between a 50w (170w total) and a 200w rider (320w total) is nearly a factor of 2. Actually, the tolerance on the calculation would be +/- 44% if they got their average right.

In case of pedantics, I know that's simplified. In practice the 200w pedaller wouldn't be averaging the same speed as the 50w one, but that's beside the point.