Hub or Derailleur, Which Is Best?

[Wisper Bikes]

I have just finished speaking to Shimano in Holland, they are not certain of the situation regarding the use of Nexus or Alfine with crank drives but have promised to come back to me next week. I then spoke to support at Madison. The chap there said that typically Shimano were very fussy about how their hubs were used and he would think there would be an issue if such a lot of torque was put through them. So still inconclusive but I will report as soon as I have heard from QC in Holland.

 

[flecc]

Thanks David, look forward to further news. All three hub gear manufacturers have always had strict low gear torque limits, one of the reasons why they limit the size of the largest rear sprockets they produce. That will still apply, but it's never previously been a reason for any of them not to supply the crank drive market, something both Shimano and SRAM have always done, particularly since 2001.

 

[Wisper Bikes]

I hope there will not be an issue Flecc, as I do fancy using the Alfine again, maybe even the 11 on this project. However I would not feel easy making and selling bikes without the benefit of a manufacturers guaranty.

As soon as I have news I will post.

All the best

David

 

[Wisper Bikes]

Just had a call back from Shimano.

The position is that their warranty would cover their hub drives on the condition that no more than 40Nm (about 500W) was exerted on the sprocket at the hub.

They went onto say that;

A, Bosch reduced the max torque on their motor from 50Nm to 40Nm to stay within their parameters.
B, So far there has not been an incidence of breakage as far as they know due to the load from an electric bike motor.
C, The most likely point at which the extra power produced by electric assistance could cause damage would be during gear change, especially from 1st to 2nd. There is however a mechanism in their hub gears that prevents gear change when the hub gears are under too much load.
D, They would be happy to see their hub gears tested on a more powerful e-bike and believe that their hubs would handle the extra torque from 45km bikes well.

It is my opinion that the guys at Shimano are very confident in their product and I now feel that if there ever was an issue they would look at it very carefully before throwing out a warranty claim.

The claims made by the chap on the NuVinci stand were, strictly speaking correct. When you add max torque of say 40Nm to the rider's input the torque applied would certainly exceed Shimano's
guidelines.

 

[flecc]

That takes care of all current legal crank drives then David, judging by reports only one matches the Bosch and the others are all somewhat below. As I mentioned above, there's always been the torque limit, which of course is most likely to be breached in the lowest gears.

One safety factor on some crank units like the Panasonic ones is the fact that very many owners reduce the rear sprocket size to raise gearing and get assist power a little above the normal limit. This automatically reduces the amount of applied torque to the hub gear.

P.S. Come to think of it, after Bosch first released their unit there were complaints that it was destroying chains in only three months and mods were expected, so it's likely the torque reduction was also in response to that.
.

 

[Wisper Bikes]

Thanks Flecc,

I can't help thinking that with all the problems created by 600W+ being put through a system (bicycle chain drive) that has only been designed to take 500W or so from a super fit rider, isn't a hub drive more sensible as far as wear and tear on bicycle components is concerned?

 

[flecc]

I can't help thinking that with all the problems created by 600W+ being put through a system (bicycle chain drive) that has only been designed to take 500W or so from a super fit rider, isn't a hub drive more sensible as far as wear and tear on bicycle components is concerned?

In many circumstances it is David. I've regularly made the point in here that if the hils to be tackled mainly don't exceed 10%, or for a fairly fit rider 12%, a hub motor will do the job perfectly and often at higher average speeds. And of course much of the country and riding territory does fit that profile.

It's when the hills get very steep that for many riders hub motors can't measure up, and that's when a crank drive through the gears will do the job. I'd hate to drive a car that only had a top gear in a steep hill area! And of course the unitary nature of the best crank systems means none of the common water and corrosion problems, tending to balance out the additional transmission maintenance needed.

All in all I think it comes down to choosing the right bike for the job, and for about two thirds in Britain I'd think the hub motor a good option. In the flat Netherlands they've already highlighted this, hub motors being by a big margin the choice there now.

 

[JuicyBike]

Perhaps the weak link is the chain...

Seriously! Perhaps a simple solution to your worries is a stainless steel chain? ...like ours...

 

[flecc]

It's not a corrosion problem though, more increased wear of chain and sprockets with the additional transmitted power Bob, plus the possible threat to hub gears. However I think the issue has been exaggerated by the initial Bosch release which was unnecessarily over-powered for a drive through the gears system. Now that's corrected I doubt there is a problem, just a slightly higher frequency of chain and sprocket change.

 

[shemozzle999]

Maybe the future is with IVT?

Automatic Bicycle Transmission (IVT) - YouTube

 

[flecc]

I don't understand what he means by a ratcheting system, and the IVT mention isn't really true, it's another CVT with a neutral lower end of range, that meaning a slipping drive transfer. It appears to have rather low power handing, he quotes 1/3rd HP, but seeing him stop it with a finger on a close radius indicates only a tiny fraction of that. A lot more information is needed before it can be considered a contender.

 

[shemozzle999]

As he said it was his first attempt. Whatever is going on inside seems to work in principle.
I agree it does not appear to function that well power wise but I can only assume the 1/3 horsepower is
a goal he is trying to achieve.

 

[Wisper Bikes]

P.S. Come to think of it, after Bosch first released their unit there were complaints that it was destroying chains in only three months and mods were expected, so it's likely the torque reduction was also in response to that.
.

Thanks Flecc, do you know if that was on derailiure type gears only or on hubs too? I believe hub chains are usually more "chunky" and strong than derailiure chains and less likely to "stretch".

 

[Wisper Bikes]

Perhaps the weak link is the chain...

Seriously! Perhaps a simple solution to your worries is a stainless steel chain? ...like ours...

Hi Bob, I had heard that stainless chains tend to be more brittle, of course they stay nice and shiny though! If you look at high end mountain bike chains that need to take massive forces, they are not usually SS. We were warned off using stainless chains but maybe we should re-visit?

All the best

David

 

[flecc]

Thanks Flecc, do you know if that was on derailiure type gears only or on hubs too? I believe hub chains are usually more "chunky" and strong than derailiure chains and less likely to "stretch".

I don't know David, but most of the bikes originally using the Bosch unit were derailleur geared.

Derailleur chains are narrower so the sprockets, and by transfer the chains, are more subject to wear. The chain roller/sprocket widths are:

Hub and Single gear = 1/8" chain roller/sprocket width

Up to 8 speed derailleur = 3.32" chain roller/sprocket width

10/11 speed vary by maker, so for all derailleurs it's best to measure the overall width:

3-8 speed - 7.2mm (all brands)
9 speed - 6.8mm (all brands)
10 speed - 6.2mm (Shimano)
10 speed (narrow) - 5.9mm (Campagnolo)
11 speed - 5.5mm (Campagnolo)

 

[Wisper Bikes]

The font of all knowledge!

Thanks a million AGAIN Flecc!

I have also been researching the advantages and disadvantages of stainless steel chains, but am getting a mixed bag of reactions with the majority seeming to favour high end stain resistant chains. Do you have a preference?

Best regards

David

 

[flecc]

I use good quality standard chains David, with no regard for rust prevention or appearance. Any bicycle chain once lubricated soon turns black and doesn't suffer from corrosion, so I see little point in fancy finishes or materials.

The only thing that might improve chain life would be the use of dissimilar materials at all wear points since steel on steel is bad engineering practice. Unfortunately that's not practical within the confines of a chain's dimensions. In any case, since standard good quality new chains and sprockets are 98/99% efficient, there's little to be gained from change.

 

[Wisper Bikes]

Many thanks again Flecc, that's roughly what I had gleaned, there are some very fancy high end chains for derailiure bikes. The features and benefits seem to be built around weight and the shape of the chain components. The emphasis was on reducing noise and making gear changes more slick when used as a set with deraileurs and cassettes chosen from the same range.

 

[greyhound_dog_1]

I always wondered if it would be feasible to have a bike with no direct link whatsoever, like these hybrid cars. Have the riders effort sent to the motor (or another motor) via electrical connection and have the rider pedalling a dynamo. Motors and dynamos are generally fairly efficient creatures so maybe it would work? Not sure what that would feel like though, guess would still need some way of varying the load you feel through the pedals.

 

[flecc]

It has been tried a number of times, most recently a university effort a couple of years ago, but nothing heard if that since. It will never really be viable due to the losses at each energy conversion stage from mechanical to electrical and then electrical back to mechanical. Even if the systems were near perfect that would probably total 20% efficiency loss or more, compare that with good condition chain and sprockets at 98/99% efficient, almost lossless.