The Ideal for a moderate-powered ebike

[Reid Welch]

Yes, Leg power is springy--it's the vital shock-buffer.

A motor, particularly a gear motor, represents a solid wall of power.
If we shift a dog-clutch transmission under so much unrelenting torque as will be supplied by a full-on gear motor (that's the worst scenario),

something's going to break in short order.

Even if the motor is only partially powered up, where's the shock buffer?
We'd need a slip clutch at the very least. Or a chain under spring loaded idler arm tension.
That'd be the best kluge, but still a kluge, because discrete shifting transmissions driven by hard motive sources need some sort of declutching
or buffering as per a chain tensioner on the power side.

my opinion.

 

[Matt Gruber]

Reid
if u had the right motor, perhaps overvolted, speeds up to about 30 mph, shifting is not needed or helpful.
My 36v 500w 1020 hums along at 52v and 24.8 mph. there is NOWHERE that i want to shift.
Electric motors have a built in low gear, on a hill they work best at 60-80% speed. It's only if it drops below ~60% that a lower gear is helpful. so if 20 is normal, 12 is on a hill. do u really want a lower gear to go even slower? Not in Miami.

 

[safe]

The folks at Rohloff explained this very well:

...seems that the reason that they "overbuild" the bike hubs is that "peak" torque while pedaling can be VERY HIGH because it's a cyclical behavior. (the torque rises and falls with each revolution) Since it's not "continuous" but "peak" torque that breaks parts the designers are forced to deal with the strongest possible man at his highest possible torque. The electric motor is "smooth" in comparison.

So it is "safe" to consider gearing if YOU KNOW YOUR MATH... :lol:

(you have to "do the numbers" before you try anything)

However, people that pedal AND use the motor are going to have to be even more careful with the base gear ratio. You MUST run a tall gear for first gear or you run the risk of adding too much of a load to the hub. The "pure electric" (no pedals) machine can actually afford to run lower gears if it wants because the peaks don't occur like with pedaling.

So in the case of the Rohloff you would want to run LESS than the stated 100 Nm at the hub in order to protect against the added input from pedal power.

But here's where things get interesting... the powerband of a human is mostly from 40 rpm to 100 rpm. Outside those "extremes" there is little power being produced. However... the TORQUE that can be produced at near zero rpm's can be quite large. Power is the product of torque with time. So you can eliminate most of the problems if you can be certain that the motor is not at full at the same time that the bike is stationary. It's that "first pedal" off the line (the dragster surge) that will blow up your hub. Once moving things go way down...

The "bottom line" on all this is that if you are using a "pure motor / no pedal" solution you can go way up to around 2000 Watts before you run a risk of failure. But if you are using a "hybrid" of pedaling and motor power you need to keep the motor power way down (like 746 Watts) and keep the base gearing pretty high. But this is logical to do anyway because you would want to gear for "top speed" with the small motor and then use the pedaling to assist on the hills. So pick one "concept" or the other, but DON'T try to do both...

 

[Matt Gruber]

safe
post the
SHOCK LOADS OF A FREEWHEEL
TOO HIGH SHOCKS KILL ANY TRANS

 

[fechter]

On some of my motorcycles, the rear sprocked was mounted on rubber shock mounts in the hub. On many ICE cars, the clutch disc has springs between the disc and the output shaft. I'm not sure how you could implement something like this on a bike, but it's certianly possible.

A spring loaded idler would be one way.

 

[safe]

SHOCK LOADS OF A FREEWHEEL

Pedaling produces a torque curve that cycles up and down.

A good analogy is with what is known as a "Rogue Wave" in the ocean. Most of the time you get small, predictable waves, but every once and a while the wave size becomes HUGE and some big ships have been hit by them. (like 100 feet of water)

So it's better to use an electric motor ALONE because it's very predictable. The combination of a constant torque (from the motor) and a wildly variable torque (from the pedals) is what creates a problem.

I've been running a Shimano freewheel for about 400 miles and so far it's been fiine. I'm also careful to do "clean shifts" and back off the power when shifting. My "numbers" show that I'm putting OVER 100 Nm at the hub (around 120 Nm) when I'm using my lower front sprocket (14 tooth) and am on my first gear in back. (28 tooth)

 

[safe]

'Opposing Torque' Motor?

Here's a wild idea...

What about an electric motor that was designed so that the pedal cadence (at the point of maximum torque) would automatically halt the motor torque?

The idea would be to smooth out the power delivery to the rear wheel and the electric motor would "fill in the weak spots" in between the pedal power peaks.

With such an idea the human/electric interface would be perfected so that as the human added more the electric motor added less... the net output would be a constant and therefore you could lower your overall transmission worries. This would really help if you got into the 2hp power range because you are getting close to the limits of "normal" bicycle capabilities. (peak power for a human for short bursts is about 2 hp)

 

[knightmb]

'Opposing Torque' Motor?

Here's a wild idea...

What about an electric motor that was designed so that the pedal cadence (at the point of maximum torque) would automatically halt the motor torque?

The idea would be to smooth out the power delivery to the rear wheel and the electric motor would "fill in the weak spots" in between the pedal power peaks.

With such an idea the human/electric interface would be perfected so that as the human added more the electric motor added less... the net output would be a constant and therefore you could lower your overall transmission worries.

I see the advantage of smoothing out the power usage, but wouldn't it cause the fist hill you encounter to only have as much "climbing" power as the rider instead of a combination of both?

This would really help if you got into the 2hp power range because you are getting close to the limits of "normal" bicycle capabilities. (peak power for a human for short bursts is about 2 hp)

I wonder what the peak hp is for, say a horse? Just something I've always wondered since we still use hp over here in the states.

 

[xyster]

So it's better to use an electric motor ALONE because it's very predictable.

I was thinking last night after reading his posts, Reid's position on this is all washed up. Because peak torque is much higher, pedaling at an average of 400 watts, for instance, is going to put much more stress on the hub than the continuous 400 watt input from a motor.

 

[safe]

I see the advantage of smoothing out the power usage, but wouldn't it cause the first hill you encounter to only have as much "climbing" power as the rider instead of a combination of both?

I think you later realized this already, but I'll repeat it.

The rider that produces an "average" of 400 Watts is really pushing a torque level of about 800 Watts at "peak" and next to zero Watts at bottom. If the motor could "fill in the gaps" then the rider would produce a "predictible" 800 Watts continuous output as he pedaled or a 800 Watts continuous output if he did not. The difference would be that the motor would use less energy if the rider helped out, but you would not see any change in torque at the hub. The way it is now the pedal power would produce a torque equivalent of 1600 Watts at peak and 800 Watts at bottom... very dangerous...

It might feel discouraging that the harder you pedaled the less the motor helped out... sort of like slipping into quick sand...

 

[knightmb]

I see the advantage of smoothing out the power usage, but wouldn't it cause the first hill you encounter to only have as much "climbing" power as the rider instead of a combination of both?

I think you later realized this already, but I'll repeat it.

The rider that produces an "average" of 400 Watts is really pushing a torque level of about 800 Watts at "peak" and next to zero Watts at bottom. If the motor could "fill in the gaps" then the rider would produce a "predictible" 800 Watts continuous output as he pedaled or a 800 Watts continuous output if he did not. The difference would be that the motor would use less energy if the rider helped out, but you would not see any change in torque at the hub. The way it is now the pedal power would produce a torque equivalent of 1600 Watts at peak and 800 Watts at bottom... very dangerous...

It might feel discouraging that the harder you pedaled the less the motor helped out... sort of like slipping into quick sand...

So you are saying that each "leg" is producing 800 or 400 watts in one rotation? I'm having trouble understanding, since the peak would only be "one" leg pushing down at a time. So a single human leg can produce over 1hp? We might need to determine what type of bike we are talking about. Are they using pedals that attach so one leg is pushing down while another is pulling up? A classic mountain bike for example is using gravity to feed the power, so a heavier rider should in theory produce more power. It's like trying to stand up (use energy to defeat gravity) but never really standing up, instead the power is transferred to the pedals, chain, wheel, road, etc.

 

[safe]

So a single human leg can produce over 1hp?

Now you get it!

At peak torque for the instant that the crank is halfway down from the top position the crank produces a "burst" of energy that (depending on gearing) would translate to roughly equal a "continuous" 1 hp motor.

 

[knightmb]

So a single human leg can produce over 1hp?

Now you get it!

At peak torque for the instant that the crank is halfway down from the top position the crank produces a "burst" of energy that (depending on gearing) would translate to roughly equal a "continuous" 1 hp motor.

That's where I'm confused. One horse power is 33,000 lb.-ft. per minute. A healthy human can only do about 1/10th of that, with both legs. We know it as a much easier 745 watts here.

Using Lance Armstrong as a benchmark, he hits 600 watts max under lab conditions. Which means if they didn't have him held down he would jump off his bike and this is only at an extreme that is a very short burst. His average power riding power is already twice that of a normal human at 245 to 280 watts.

The hp power measure is based off of horses, which last I looked and rode one are a heck of lot stronger than we are. As much as I would love to make humans as strong as a horse, I just don't think we can. If even the world's top cycle athletes strongest is probably average for a weak horse and the weak horse could sustain this "power" much longer.

I still think our electric motors are way more powerful than we want to admit. If we really had 1 hp or 745 watt power in our legs, we wouldn't need electric motors to go 30+ MPH down the road on the bike.

 

[safe]

Kill PWM?

The biggest threat to the hub is the effects of PWM. At the very places you LEAST want torque from the perspective of the hub (very low rpms) you get a big surge of "extra" where you don't want it. This "extra" is a bonus for "hub motors" but comes as a negative for geared bikes because it amplifies torque in a narrow area. It's the combination of the effects of PWN and the pedal "peak" that throws the torque peak up really high in certain conditions.

Remove PWM and you virtually eliminate the problem of excess torque.

 

[safe]

Using Lance Armstrong as a benchmark, he hits 600 watts max under lab conditions.

600 Watts "continuous". (they are averaging over time to get that 600 Watts value)

Instantaneous peak will be double that...

 

[knightmb]

Using Lance Armstrong as a benchmark, he hits 600 watts max under lab conditions.

600 Watts "continuous". (they are averaging over time to get that 600 Watts value)

Instantaneous peak will be double that...

Maybe some numbers would help me understand, is it possible to plug in some "estimates" in the graph using Lance stats?

 

[safe]

Looks like "about half" is "close enough".

 

[Lowell]

I like the idea of a heavily built hub, but the fact remains that they were only designed for pedal power. Breaking a $1000+ hub would be an expensive experiment.

Just do the math. If you know what is "safe" then the hub will never break on you. Every time you lower the gear ratio you get a corresponding increase in torque. The Rohloff hub allows 100 Newton Meters at the hub and you will find that it's not that hard to keep things within that. Even with a 2000 Watt motor if the gearing is high enough you will never exceed the spec.

Know your math... save $1000... :wink:

Also, as rpm's rise the torque drops a great deal, so you can protect yourself if you make your "power peak" and "efficiency peak" be the same... translation... run a low current limit controller. Since you have plenty of gears (torque options) you can afford to run "perfectly" all the time... this also reduces heating...

The simulator predicts 135-140 Nm for my setup right now, and it could still use more.

http://www.electricrider.com/custom/index.htm

0-40mph in under 6 seconds would be acceptable performance.

 

[Mathurin]

1hp = What a mine pony should be able to give over a 4h work shift, +50% to make it a horse. An actual horse wouldn't be able to sustain this power all day, and horses have been measured up to 14.9hp for a few secs. Says wiki.

 

[Reid Welch]

So it's better to use an electric motor ALONE because it's very predictable.

I was thinking last night after reading his posts, Reid's position on this is all washed up. Because peak torque is much higher, pedaling at an average of 400 watts, for instance, is going to put much more stress on the hub than the continuous 400 watt input from a motor.

Nope, I stand by my PreDicTion 100%.

You, you are following safe's apologia, are also 100% wrong.

No geared bike hub will withstand shifting under the torque of any electric motor drive


unless that drive is torque-buffered by, uh, rubber band drive or the like.

That's the fact, heh! Prove me wrong.

Safe, you don't appreciate the POWER of what in effect is a hammer tap...tiny?
Hell, it's HUGE. It's shattering. Those internal geared hub gears and dogs are hardened steel, btw. A tough steel deeply cased.
You know what that means, right?

It wont work.

The NuVinci -could-!

:lol:

* i love playing curmudgeon today. easy though becaws I'm write!*

i am!

 

[xyster]

You, you are following safe's apologia, are also 100% wrong.

Occasionally Safe's right, and makes his point understandable and concise. There was just a lunar eclipse, maybe that has something to do with his timing.

 

[safe]

No geared bike hub will withstand shifting under the torque of any electric motor drive unless...

Think logically and not emotionally for a second. I know you want to feel attached to "high tech" and that means NuVinvi, but let's reason this out a little.

Pedaling is...... jerky.... in that you get pulse, pulse, pulse of double the average torque as it goes through the cycles.

Motors are smooth and continuous in their delivery of power.

Now as long as you back off the throttle when you shift (which I assume we are all assuming is our baseline) then the power to the hub will not have great spikes in it like pedal power does.

So if the calculated torque is properly respected with the smooth continuous torque of a motor then what is the problem? (let's leave out hybrid pedal/motor bikes that combine the two because those are more complicated... strictly motor and hub... nothing more.. no pedals)

Be logical now... no vain attempts to glorify "high tech" for "high tech's" sake....

 

[Matt Gruber]

1st u need a throttle like mine.
when u can go OFF to GO without hearing the freewheel BANG, then it could work. I've never seen any stock throttle do this. Even mine bangs if i am not careful.

 

[safe]

1st u need a throttle like mine.

These throttles aren't great I'll agree with you on that. What I do is snap the throttle closed when I'm shifting and then roll it back open again. I do it really fast so that it's almost like my own Pulse Width Modulation in that the result (if you do it fast enough) is that the motor just slightly backs off enough for the shift. It works okay, the freewheel has held up to everything so far. (almost to 500 miles now)

Just went out for another ride... made it to 50 mph somewhere else... so I've done 50 mph in more than one place now... could not beat it though. (I think I need to find a "real" hill... which in the midwest is not easy to do) The handling is great, but I really need a fairing. I can feel how much air is getting trapped right in front of my face and if I could get a fairing it would really improve things a lot. Later this spring/summer is the soonest I'll get to it though.

 

[safe]

Kill PWM?

I'm reposting this because it got lost in the shuffle...

The biggest threat to the hub is the effects of PWM. At the very places you LEAST want torque from the perspective of the hub (very low rpms) you get a big surge of "extra" where you don't want it. This "extra" is a bonus for "hub motors" but comes as a negative for geared bikes because it amplifies torque in a narrow area (the low efficiency heat producing area) that the geared bike shouldn't ever need or want. It's the combination of the effects of PWM and the pedal "peak" that throws the torque peak up really high in certain conditions.

Remove PWM and you virtually eliminate the problem of excess torque.