Belt Drives; Better for E-bikes? Gates has one.

[Nehmo]

E-bikes are clean. But there’s one part that’s not so clean: the greasy chain and chain related stuff. Why not get rid of it? A good E-bike should only have the pedals for contingencies. The drive train and transmission on an e-bike can be reduced to a clean, minimal-maintenance belt and two accommodating pulley. Gates has a system http://www.carbondrivesystems.com/forbikemakers.php?lang=us (more info http://www.carbondrivesystems.com/downloads/tech_docs/GATES_TECH%20MANUAL_ENG-CHI_REV02_REL21.pdf ). You can limit the bike to one gear ratio or you can get an internally geared hub. There’s even a $100 2-gear ratio hub that doesn’t use cables Sturmey-Archer S2 Kick-Shift Hub Kit
http://road.cc/content/review/33373-sturmey-archer-s2-kickshift-hub. But hubs deserve a discussion of their own.

Newly designed e-bikes should incorporate a belt drive system.

But this makes for one new problem. The chain stay part of the frame will need to be breakable.

Anybody see any other problems with belts?

 

[Hugues]

i'm always in to reduce complexity,

but don't belts generate more friction than a nicely lubricated chain ? not sure...

but for me, i ride in the mountains, yesterday i reached the end of my battery and i was REALLY glad to shift down from 27 to 1st gear to go back home.

but then i don't need 27 speed, i guess 2, 3 or 4 would be enough.

 

[John in CR]

I've run out of gas the same number of times I've run out of battery juice...twice. I have no plans to repeat either event.

 

[Demosthenes]

For an ebike that is not being pedaled they would work fantastic. But I disagree that pedals are merely a contingency measure. I myself pedal all the time. Here is a quote from Sheldon Brown's website. An opinion I certainly trust.

"From time to time toothed belts are proposed as an alternative to chain drive. The major advantage touted for this system is that it is cleaner than a chain.

There are serious disadvantages to this approach, however:

Belts can't be opened and re-closed as chains can, so it is not possible to fit one to a standard diamond frame, unless you cut the frame open, because the belt has to be looped through the triangle formed by the right chain stay, right seat stay and seat tube.
Belts can't be lengthened/shortened as chains can, making it difficult to customize the gearing by going to a different sprocket/chain ring size.
Belts bend instead of pivoting, and are less efficient than chains.
Belts are stretchy. To avoid skipping, they have to be kept tensioned, further reducing efficiency.
Belts can't be made as narrow as chains, so they do not lend themselves to derailer gearing."

So to somebody who doesn't pedal. 4 out of 5 of those can be ignored. Leaving only the one you mentioned. Still, there's another perspective for you.

 

[Nehmo]

i'm always in to reduce complexity,
but don't belts generate more friction than a nicely lubricated chain ?

I advocate a cog-like belt system – not an engagement-by-friction system. In the Gates system, which is cog-like, they say 3rd party testing reveals their system is as efficient as a chain. It looks as though the friction loses would be minimal.

but for me, i ride in the mountains, yesterday i reached the end of my battery and i was REALLY glad to shift down from 27 to 1st gear to go back home.

That’s the contingency implied above.

but then i don't need 27 speed, i guess 2, 3 or 4 would be enough.

Those 27 possible gears represent only about 20 distinct (different) gears and 15 practical (big front to small back is too much of an angle, etc.). You do have a significant gear range (proportional difference between lowest and highest), possibly 500%.
The number of steps needed to cover the range is one issue. The size of the range is another. But the importance of anything related the gears to transmit human power is low if the e-bike is electrically powered for the whole trip.
If the e-bike is reliable, a belt drive and a 3-gear ratio hub with a gear range of 160% is more than necessary. http://www.amazon.com/Sturmey-Archer-S3x-Fixed-Gear/dp/B0042R50QQ
BTW, you certainly can solve the derailleur problem by using a $1,200 Rohloff Speedhub, 14 speed with a 526% range.

 

[Nehmo]

Here is a quote from Sheldon Brown's website. An opinion I certainly trust.

In some fields, experts aren't really. If SB is so cool, where are his e-bike comments? Isn't this a major development with bikes?

"From time to time toothed belts are proposed as an alternative to chain drive. The major advantage touted for this system is that it is cleaner than a chain.
There are serious disadvantages to this approach, however:
Belts can't be opened and re-closed as chains can, so it is not possible to fit one to a standard diamond frame,

Yes, the frames have to be designed accordingly, and this represents something a little different. But frames are currently designed for derailleur & chains too.

Belts can't be lengthened/shortened as chains can, making it difficult to customize the gearing by going to a different sprocket/chain ring size

I'm not advocating using the belt as a drive train & a transmission - just the drive train.

Belts bend instead of pivoting, and are less efficient than chains.

There's not much difference in effiecency.

Belts are stretchy. To avoid skipping, they have to be kept tensioned, further reducing efficiency.

That's not true with some belts, and I wouldn't advocate using a "stretchy" or elastic belt. Indeed, Gates points out the inability for the Carbon Drive belt to grow as a design constraint. Frames with rear suspension can't have the pivot point behind the bottom bracket because this requires belt growth, which the belt cannot do.

Belts can't be made as narrow as chains, so they do not lend themselves to derailer gearing."

Again, the derailleur needs to die. But the width of the Gates belt is 11 mm.

So to somebody who doesn't pedal. 4 out of 5 of those can be ignored. Leaving only the one you mentioned. Still, there's another perspective for you.

Well, I do peddle, now anyhow. But I'm designing a bike that will rarely require it.

Typically the electric drive is retro-fitted on an existing bike, and these already have the ubiquitous derailleurs-with-chain. It's time to design from the beginning focusing on e-power.

 

[Nehmo]

I've run out of gas the same number of times I've run out of battery juice...twice.

You may regard yourself as a lucky man with forethought.

 

[Demosthenes]

Here is a quote from Sheldon Brown's website. An opinion I certainly trust.

In some fields, experts aren't really. If SB is so cool, where are his e-bike comments? Isn't this a major development with bikes?

Well unfortunately he did pass away about three years ago. Admittedly his opinions aren't entirely relevant to ebikes.

You'll probably never convince those who prefer to use the motor as a secondary to pedals source of power but to those who it is primary then I totally agree. Some belt drive ebike specific frames and bikes would be an amazing thing to have. I was putting my bike into my girlfriends car the other day and would have loved a belt drive, we got absolutely covered in grease from the chain.

http://www.norco.com/bikes/urban/performance-urban/ceres/
http://www.norco.com/bikes/mountain/29ers/judan-belt/
Norco has a couple belt drive bikes if your curious. I've tried the one at work. It's definitely intriguing tech.

 

[Hillhater]

Sheldon Brown wrote:.....Belts are stretchy. To avoid skipping, they have to be kept tensioned, further reducing efficiency.

That's not true with some belts, and I wouldn't advocate using a "stretchy" or elastic belt. Indeed, Gates points out the inability for the Carbon Drive belt to grow as a design constraint. Frames with rear suspension can't have the pivot point behind the bottom bracket because this requires belt growth, which the belt cannot do.

If you have ever worked with toothed power transmission belts ( as opposed to "timing" belts).. you will know that they need to be tensioned to ensure secure drive,..Its not because they stretch, but to prevent "skip" on the pulleys.
High torque drives demand a LOT of tension.

 

[amberwolf]

It's no different from chain drives in that respect. If you have a loose enough chain, and a small enough sprocket engagement region (especially on the highest gears on rear cassettes), it'll skip, too. Used to happen a lot to me on CrazyBike2's original powerchair motor drivetrain.

 

[Hillhater]

It's no different from chain drives in that respect. .

.. Chain drives should not need any actual "tension" pre-load, and are normally set up with a minimal amount of "slack" movement in the return side. Over tensioning results in rapid wear of chain and sprockets.
The other issue with belt drives is the importance of maintaining precise alignment to prevent riding off the pulleys.
Chain alignment is important for wear and efficiency, but by their nature,they can tolerate a lot of miss-alignment ( eg a derailleur drive)

 

[Nehmo]

If you have ever worked with toothed power transmission belts ( as opposed to "timing" belts).. you will know that they need to be tensioned to ensure secure drive,..Its not because they stretch, but to prevent "skip" on the pulleys.
High torque drives demand a LOT of tension.

I'm not sure what you mean by "High" and "a Lot". And being "stretchy" (Sheldon Brown's word) and requiring tension are different things.
Sure, a toothed drive belt would require a particular tension to function properly. Are you saying this amount of tension would be so high that the frictions loses would make the efficiency of the system unacceptable?

 

[number1cruncher]

I have a Gates belt drive on my GT. The bike in my avatar. Its ok, but does have its pitfalls.

First, it likes to jump the track occasionally.
Second, changing the tire is a nightmare.
Lastly, I have not been able to source a single new belt. Plenty of kits with all the gears and crap... I've called GT, I've called Gates and they all act like I'm asking for something they don't sell or build.

As for the frame: My GT has an elongated rear triangle, so the belt stays above the chain stay. I have seen a couple break-away frames, but they did not look durable.

 

[grindz145]

I've run out of gas the same number of times I've run out of battery juice...twice.

You may regard yourself as a lucky man with forethought.

I've ran out of gas many many many more times on my motorbike than my ebike, in fact, my ebike has better range as well.


P.S. belt drives can actually handle the torque from ebike motors, bicycle chains are not equipped. Belts last longer if designed properly, apples to apples, but they are prone to collecting road crap (stones and stuff) which would make them not the first choice for off-road applications.

 

[Nehmo]

I have a Gates belt drive on my GT. ...
As for the frame: My GT has an elongated rear triangle, so the belt stays above the chain stay...

I can't make a mental pic of that even with looking at your avatar. What do you mean? Do you have a pic posted somewhere?
And how recent is your belt? I'm considering the latest version (http://www.biking.com/buy-bike-accessories/gates-carbon-drive.html). I plan on modifying an existing frame.

 

[Nehmo]

@number1cruncher, I found a pic of your bike. The rear end is hard to understand from the pic, though.
http://www.gtbicycles.com/bikes/urban/comfort/2011-aerostream-chocolate

In writing these posts, I've done some thinking. I believe an e-bike needs to have a rear suspension, and if this Gates Carbon belt is to be used, then a bottom bracket centric pivot (BB-centric suspension) would be best
And, with vertical dropouts, an Eccentric Bottom Bracket (EBB) would be one method to tension the belt.

 

[Miles]

A good E-bike should only have the pedals for contingencies.

Well, there's a value judgement masquerading as a fact....

 

[number1cruncher]

Here's a picture of the ladies version of the GT Aerostream:



Look at the rear triangle and how it is below the belt near the dropout.

 

[spinningmagnets]

As long as we're on the subject, there are rear-suspension frame styles with no chainstay

Also, heres one example of a chain-stay splitable joint (bolted during use), and a bolted-on drop-out reinforcement that covers a slot in the frame.

 

[Hillhater]

Am i alone in thinking that this is a very poor design for a eccentric BB ?
It seems that much of the frame rigidity and strength depends on those two 6mm bolts !

 

[spinningmagnets]

I hadn't thought of that. I think you're right. The bolts in that location are under tension, and the joint would be better under compression, perhaps between the down-tube and seat-tube (at roughly 11:00 O-Clock, rather than 6:00)?

 

[GCinDC]

my minibike 30 yrs ago had a belt and a centrifugal clutch. ive been wondering if that would work on an ebike. any thoughts?

 

[Hillhater]

I hadn't thought of that. I think you're right. The bolts in that location are under tension, and the joint would be better under compression, perhaps between the down-tube and seat-tube (at roughly 11:00 O-Clock, rather than 6:00)?

Yep ..exactly !
..I sometimes wonder at the thought train ( or lack of !) on some bike designs.

 

[amberwolf]

I agree, but FWIW all the pics of EBBs I've ever seen were split the same place--the bottom of the BB shell. I'm sure some aren't, but I've not seen them. Sometimes the tensioning mechanism is different, but the split was the same point.

However, a bit more thought and I wonder....since the toptube and seatstays are acting in compression to take the loads that the downtube and chainstays are taking in tension, and the seattube directly connects all these points, vertically thru the (normal) main load/weight, does it really matter that the BB shell is split at the bottom?

I mean, if you just cut the shell of a BB itself on any bike, and left it that way, not using it for pedalling, would anything break because of that? You'd want to do it on an ebike so you wouldn't have to even have the cranks in there (or have them but not use them, since using them *would* force the shell more open), but would it actually cause any problems?

 

[gtadmin]

Am i alone in thinking that this is a very poor design for a eccentric BB ?
It seems that much of the frame rigidity and strength depends on those two 6mm bolts !

Would certainly agree - if there was no top tube or seat stays! However, there will not be a huge amount of force there trying to open the "split", and then there are two 6 dia bolts holding it together. Work out how much force it would take to stretch both of them (and without reverting to any formula, gut feel says a couple of kN)

GT