Shimano's patent for a new gearbox design.

[Randy Butternubs]

Looks interesting, but if you are going for a more expensive, complicated, heavy system than a normal derailleur then I wonder if it wouldn't be easier to make a system that integrates a derailleur with a sealed, oil-bath chaincase. Especially if you are designing the frame around it anyway.

I do love the idea of a sealed drivetrain but current offerings are too pricey for me or have other issues. As above, I'm not convinced that lubrication or chainline are very important.

 

[neil_merseyside]

Why not just a rohloff in a "chain case" using a belt.

 

[overmind]

Why not just a rohloff in a "chain case" using a belt.

Agreed. It seems to me that the easiest/best/simplest place to have any kind of gearbox design is in the hub of the back wheel (a la Sturmey Archer, Shimano Alfine/Nexus, Rohloff). Then just have a single chainring on the front with a belt drive.

The bottom bracket area then remains unchanged (instead of looking like it has a colostomy bag attached).
When the galaxy hub wears out just replace the wheel (or rebuild the wheel with a new hub). It is a lot simpler.

Perhaps the gearbox on the front allows a larger range of gears. This is the only reason I can think it might be needed.

(Speculation) I wonder if you could have a much simpler version of the hub gears but on the front (as well as the back). This would allow a much simpler/smaller mechanism at the bottom bracket end of the chain.

(I just checked)
LOOK. It already exists:
http://www.schlumpfdrive.com/index.php/speed-drive.html

If you combined the schlumpfdrive with a Rohloff or Alfine/Nexus hub you could have a huge range of gears (3 X 11) You could then either use front/rear shifters in the traditional way or have some kind of electronic gizmo (using stepper motors and such) with a single shifter. I reckon you would have about 30 gears.

 

[confusedcyclist]

Agreed. It seems to me that the easiest/best/simplest place to have any kind of gearbox design is in the hub of the back wheel (a la Sturmey Archer, Shimano Alfine/Nexus, Rohloff). Then just have a single chainring on the front with a belt drive

Agreed, but unless you are running disc brakes, rims in particular are considered consumables and the cost of wheel rebuilds adds up. Removing the hub issue from the equation makes the drive train service straightforward.

I can see this being marketed to low maintenance types.

 

[Drago]

Take a Sturmey Archey 3 or 5 speed epicyclic job.

Jazz it up to accept maybe 8 or 10 carefully selected gears.

Mount the body of the hub in the bottom bracket.

Re engineer the hub so the spindle becomes the input side of the design, and mou t the cranks upon it.

Re engineer the hub so the sprocket becomes the output side, and connect to to the rear wheel with a chain.

Twiddle things a bit to the selector cable exits the body somewhere useful.

All done. No chains, liable to function for 50 or a hundred years with simple maintenance.

 

[classic33]

Take a Sturmey Archey 3 or 5 speed epicyclic job.

Jazz it up to accept maybe 8 or 10 carefully selected gears.

Mount the body of the hub in the bottom bracket.

Re engineer the hub so the spindle becomes the input side of the design, and mou t the cranks upon it.

Re engineer the hub so the sprocket becomes the output side, and connect to to the rear wheel with a chain.

Twiddle things a bit to the selector cable exits the body somewhere useful.

All done. No chains, liable to function for 50 or a hundred years with simple maintenance.

Combine with a triple up front, and front and rear changers and you can have a 105 speed bike.

Edited to add the picture

 

[TheDoctor]

Pffft. I'll stick with my 7 speed bike, which I built long before "1x" became a thing.
It needs a 12-32 cassette, a 36T chainring, and a chain, every five years or so.
This looks over-designed, over-engineered, and overweight. And needs a dedicated frame.
I'm out.

 

[Tigerbiten]

If you combined the schlumpf drive with a Rohloff or Alfine/Nexus hub you could have a huge range of gears (3 X 11) You could then either use front/rear shifters in the traditional way or have some kind of electronic gizmo (using stepper motors and such) with a single shifter. I reckon you would have about 30 gears.

Been there and done that ........
I've a Schlumpf HSD with twin chainrings and a Rohloff on my bent trike.
It's a very good setup if you want an extreme gear range, mine is very close to 1900%.
But unless you need ultra low and ultra high gears beyond the range of a standard triple and cassette on the same bike, it's not worth it.

Luck ............

 

[keithmac]

After nearly 3 years on my belt drive commuter I'd be loathed to go back to a chain.

This coupled with a belt drive would be great for just wheeling out of the shed and riding to work on in all weathers!.

As said though the frame would only take one gearbox so you'd be stuck with it, I'm sure they will make it fit the STEPS equipped frames, would be daft not to..

 

[Gravity Aided]

Pinion P1 .18 gearbox and a Gates belt seem to do this job already. In production. I think Co-Motion Cycles offers these, and they build mostly touring bikes. Fairly custom stuff. Well above my paygrade. I prefer stuff I can afford, and repair. I've been tackling hills, albeit small ones, for some time with the current set up.

 

[Shadow121]

It will certainly cost more than most people are prepared to pay - initially. But that's true of all new technology, once R&D costs are recouped and production levels rise prices fall. Integrated shifters and carbon frames cost eye watering amounts when they were introduced yet now both are common currency. Shimano don't faff about when they develop new systems, it isn't guaranteed to succeed but it's certainly a case of "Watch this space".

It also makes belt drive a very doable proposition, something not possible with derellieur gears.

Gears is old technology, been around longer than this generation.

 

[MichaelW2]

Looks like it would be good for commuting

So often the technology that is best suited to commuting is shoehorned into the top end race groupsets then trickled down over the following seasons. Disk brakes are a bit meh for racing but ideal for a commuter road bike. We need to work on the concept of trickle up. Bring a useful new concept in at the 105 groupset level and if racing folk want it, they can experiment. Another benefit of trickle up is that the technology would be designed for the affordable materials of midrange rather than requiring high end metallurgy and manufacturing techniques that do not trickle down so well.

 

[MichaelW2]

Acknowledging the BikeRadar presenter's caveats, he seems to suggest that the potential 'perfect' chainline offered by such a gearbox is a key 'efficiency' factor. Not sure about that.

Conventional attitudes to chainline assume a long chain with some lateral play built in to the chain design. This enclosed transmission requires very short chains with no room for lateral play. If you hold 2 cassettes big to small as close as possible, that is how long the chain has to be.

 

[MichaelW2]

Most hub gears have a limit for lowest gear that can damage the internal mechanism. Adding a Schlumph drive 2 ratio front end may take you beyond the limit.

 

[Ajax Bay]

Conventional attitudes to chainline assume a long chain with some lateral play built in to the chain design. This enclosed transmission requires very short chains with no room for lateral play

By "conventional attitudes" do you mean that, as the paper concludes, chain-line offset (normal length chain) has a negligible effect on efficiency? The enclosed very short chains will, I surmise, be at relatively low tensions, though (relative to the chain tension* experienced in a conventional chainring to sprocket combo). And since "efficiency varied with the reciprocal of the average chain tension [highest efficiencies occurring at high chain tensions]" this will mean losses in the dead in-line, very short chain runs, far outweighing the minimal loss of chain-line offset.
Edit to add: detail: https://www.cyclechat.net/threads/shimanos-patent-for-a-new-gearbox-design.255042/#post-5795536
*About 200N at 200W power (big ring circumference 75cm, cadence 80rpm) and maybe 300N if in the small ring (same power). (Please say if my maths is way out.)