Bicycle Mechanics - Thoughts on the upcoming SRAM 9-speed hub (and the future of internal gears)?

Yesterday I was goofing around with Sheldon Brown's Internal Gear Calculator (http://sheldonbrown.com/gears/internal.html) when I noticed a new option available - the SRAM i-Motion 9-speed hub. I did some searching, and there isn't too much info out on the web about this hub just yet, but I did find some interesting things.

1. A link to a PDF of specs for new SRAM technology on their website here: http://www.sram.com/en/service/sram/tech_specs.php

2. An article discussing it here (scroll to the bottom of the page): http://bicycletech.blogspot.com/

3. An MTBR thread discussing internal drivetrains in general here: http://forums.mtbr.com/showthread.php?p=1786335

4. And today, the official i-Motion website here (can anyone that speaks German give us the gist of what they're saying in the movie?): http://sram-imotion.com/

I'm not going to say much about the hub itself - I don't know much about the actual industry, so I'm in zero position to speculate about anything specific other than explicitely released information about specs, but given the specs on paper and SRAM's (both home-grown and inherited from Sachs) generally rock-solid reputation for quality components, this looks like it could be a real winner of a hub.

In my opinion, it looks like the future is bright for internal gear systems, and the next-generation hub market is really beginning to swell. It looks, to my layman, somewhat biased eyes, that derailer systems are beginning to (if they have not already) reach their practical limits for further development. There just aren't too many more cogs that we can cram onto a hub without seriously compromising wheel strength or durability. Increasing hub spacing to compensate could start producing chainline and biomechanical problems. Furthermore, adding all those cogs hasn't improved the number of useful gear combinations by a whole lot.

On the other hand, the development future for internal gear hubs looks like it might be wide open, especially in mountain biking, where rapid, easy gear shifts, strength and durability are important concerns, and in commuting and utility cycling, where low maintenance requirements and all-weather durability and performance are important. In fact, I think that the concept has long-since been proved (by Sturmey-Archer, in the first half of the 20th century) in the case of utility cycling, but the new hubs are beginning to make internal gearing more appealing for utility riders in hilly cities and for more performance-oriented, enthusiast commuters and utility cyclists. In case of mountain biking, hubs like the i-Motion 9 are beginning to reach the necessary range of gears for cross-country riding, without the chainline issues, shifting challenges and huge number of overlapping, useless gears in a 27-speed MTB drivetrain. I know that the Rohloff is already there, but the Rohloff remains a bit of a boutique product. In order to compete with derailers, the price of quality, wide-range internal gearing must come down.

Unfortunately, I really don't know what the mechanical and financial constraints on further development of relatively inexpensive wide-range internal gear hubs are. I don't know if there isn't much more than can be done, or if we're seeing something like the kind of rapid technological innovation that the derailer went through some 25 years ago, with the new Shimano and SRAM hubs being somewhat analogous to SunTour's slant parellogram - a major change, followed by a long series of refinements to the technology. I'm hoping that some of the more industry and tech-savvy BF members might be able to shed some light on this.

In any case, while it looks to me like we'll be seeing a lot more internal gear hubs in the future, both for utility and high-end use, I don't think it's very likely that the derailer will be entirely supplanted. After all, derailer gears are cheap, making them an appealing option for a wide range of gears at low cost. The only current hub that provides enough range for touring is the Rohloff, and as I said, I don't know how practical 14-and-higher-speed hubs will ever become, in terms of cost. I'm optimistic, but open to the possibility that they'll remain expensive, luxury items for the forseeable future. I also don't know if we'll ever see internal gears on fancy-pants racing bikes. This is due to weight (though the difference is less than it is generally made out to be), mechanical inefficiency (also not as bad as supposed, especially with the new hubs - I predict improvements here, too) and gear spacing that is too wide for maintaining optimum cadence at all time (a legitimate complaint against hubs, if you're a racer).

Anyway, that's my long-winded, very-very-non-expert take on the new SRAM hub, and on the future of internal gears. So... what do you think?

The only experience I have actually riding internal gear hubs was on a Japanese 3sp from the 1960 era. First gear was very sluggish and if the temp was below 5C, you were in real trouble. I have many miles on deraileur systems with no really bad experiences to report. I like the broad and easily modified gear range and the ability to provide wider jumps between lower gears and narrower jumps between higher gears. I do not find them to be high maintenance or finicky. The thing is that I have done some touring in situations where IF I had sucked a branch in a rear der. I might have faced hypothermia or worse. I also commute and have a marked disinclination to maintain that bike (or ever even wash it). Sooo I am certainly open to internal gearing. I would love a reasonably priced system with sensibly spaced gears....I think that to drive tech. dev. we need many more riders with a higher degree of technical expectation than the typical entry level cyclist. There is no doubt that this can happen...the profit margins must exist for this to become reality.

Good thoughtful essay.

The drawbacks to internal gear hubs are greater weight and lower efficiency which will probably not be solved soon, if ever, and that will keep them from replacing derailleur gearing systems for performance riders. They certainly are making inroads in the utility market and are excellent tools for that use.

I wasn't aware of the new SRAM 9-speed unit but, as you noted, Sachs was in the market in a big way in the past. Shimano has both 4 and 7-speed internal gear hubs and, of course, Rohloff has the very high end of the market to itself at present.

It will be interesting to see how this all develops.

I'll consider buying one, but looking at the specifications I'm a little disappointed by the gear range of only 340%.

The gear spacing is a little closer than on my S7, but I've toured fully loaded with it and the gear steps weren't a problem - the problem is that I need to match the sprocket I use to the terrain I'll be cycling so if I'm going somewhere very hilly my top gear only allows me to pedal on the flat at about about 32km/h, or if its mostly flat I don't have a low gear for hills.

I'd have been more likely to buy the i-9 if it had a bigger range - not sure its different enough from my S7 to justify. I don't think its quite there yet - 11 or 12 gear with a range >400% and it could be competing with the Rohloff. I'm sure there'll be another hub from SRAM in a couple of years time that'll be closer to the Rohloff, so probably I'll just wait until then.

There was a thread a little while back where some guy was trying to convince everyone that hub gears are just a marketing ploy and that they don't offer anything over derailleur systems. I absolutely disagree with that. If Rohloff hub gears were half the price they are I'm sure the big manufacturers would be specifying them on their bikes.

Now I just want to know how much the i-9 is going to cost.....my guess would be about $500 all in.

Hrm. I happen to think that 340% is a pretty respectable gear range. Remember that a typical 70's 10-speed (with 39/52 cranks) had a gear range of about 267%. If we're generous, and give it a 34-tooth freewheel cog, that 10-speed has a range of 324%. Not really suitable for touring, but pretty good for anything else, I think. No, it's not at the 500%+ range that a derailer system can manage, and I think that's a good reason to prefer derailers for some purposes. It is getting closer, though, and for most applications, 340% is as much as anyone could ever need, I think.

I'm one of the posters from the other forum and I'm glad to see a bit of discussion here also. I'm happy to see some new products coming out in the gearhub world. The Rohloff thus far is the best but it is pretty expensive for average cyclists. The nexus premium 8 speed with its ultegra level bearings and seals is a nice improvement over the older nexus 7. Hopefuly Sram's new i9 will be an improvement over the Nexus 8. Sram has gotten rid of the external click-box on the i9 so it should be a little more durable. Somepeople have complained of less than perfect shifting with the older sram 7 speeds so hopefuly they have ironed out any shifting problems. Since sram offers its dual drive with disc brakes maybe they will see fit to offer discs on the i9. Since the dual drive is basicly a three speed gearhub with a cassette mounted on the end. Maybe this will spur Shimano to offer discs on the nexus. They are all ready available aftermarket. Hopefuly both manufactors will see the light and start offering roadbike style shifters for their gearhubs. Think ultegra level 10 speed brifters. I think a gearhub on a light touring or road bike make sense for a lot of people. As of yet shimano does't recomend the nexus for offroad riding. Not sure if sram will rate the i9 durable enough for off roading. With the available sprokets and chainrings it is ez to gear a nexus8 or i9 from around 30 gear inches to 100 gear inches. Which is good enough for me. Some others may need more range. Overall I think gearhubs offer most cyclists a durable low maintenence choice and I'm glad to see some new products coming out. Hopefuly they will be improvements to what we already have.

Since the dual drive is basicly a three speed gearhub with a cassette mounted on the end.
This always impressed me as the worst of all worlds; the weight and inefficiency of an internal geared hub coupled with the complexity of a derailleur system. The only time its usage made any sense was on bikes (like some Bike Friday models) that couldn't use a front derailleur.

I have a Specialized Sirrus... what would be my out-of-pocket cost be for me to strip off all of the derailleurs (Sora, no great loss) and get a sturdy commuter wheel made with this nine-speed hub?

I guess I would need a nine-speed index shifter... is that any different than a nine-speed index shifter for a derailleur bike?

I guess I would need a nine-speed index shifter... is that any different than a nine-speed index shifter for a derailleur bike?
The shifters for internal gear hubs are completely different and incompatible with derailleur shifters. Maybe some day Shimano will make an STI-type shifter for internal hubs but not yet.

Derailleur gears are great if you keep them clean, but like some of you, I'm waiting for that one internal gear hub that clinches it for me, less expensive than a Rohloff, but with plenty of gears for hilly commutes/touring. I've got a 3-speed for neighborhood runs, and I love the simplicity. Thanks for the links and and good input, everyone. When the right hub comes along, I'll probably switch all my bikes and never look at derailleurs again.

QUOTE=HillRiderThis always impressed me as the worst of all worlds; the weight and inefficiency of an internal geared hub coupled with the complexity of a derailleur system. The only time its usage made any sense was on bikes (like some Bike Friday models) that couldn't use a front derailleur.

It eliminates the front derailleur and improves chain line. A 3 speed internalhub weighs very little. Not sure what the weight difference between 3 chainrings and a derailleur would be versus a 3 speed internal hub and really don't care. It allows the rider to drop into a low gear if they come to a complete stop with the rear derailleur in a high gear. Thus enabling them to easily ride away from their stop. Different people are impressed by different things especially if their needs are different.

[QUOTE=HillRider]

The shifters for internal gear hubs are completely different and incompatible with derailleur shifters. Maybe some day Shimano will make an STI-type shifter for internal hubs but not yet.


Sad but very true.

"The Dancing Chain" gives a good account of the evolution of epicyclic gears in Britain (due to the wet climate?) versus derailleur gears on the continent. Epicyclic gears are always going to suffer from excess weight, reduced efficiency, reduced reliability under heavy loads, and expensive repairs. Commuters, beach cruisers, and transportation cyclists have enthusiastically rediscovered them in recent years, but I will be very surprised to see them used in racing or serious club riding, as they were in the 1930s and 1940s.

One can now configure a 30-speed derailleur setup which provides up to 25 non-redundant, non-crossed gear ratios over any desired range within 20 to 130 gear-inches. That is pretty hard to beat.

[QUOTE=HillRider]
The shifters for internal gear hubs are completely different and incompatible with derailleur shifters.
... I have shifted a Sturmey-Archer 3-speed with a friction-type downtube lever originally designed for a Huret derailleur. Worked like a champ, and gave me a nice neutral position between 2nd and 3rd. :)

QUOTE=John E

One can now configure a 30-speed derailleur setup which provides up to 25 non-redundant, non-crossed gear ratios over any desired range within 20 to 130 gear-inches. That is pretty hard to beat.

That is one heck of a gear spread you have come up with, 110 inches. I guess you had to use a triple chainring for that? I didn't know they made 10 speed cassettes with that much range. Could you please tell me your chainring sizes and smallest and largest cassette cogs? Thanks.

QUOTE=John E

I have shifted a Sturmey-Archer 3-speed with a friction-type downtube lever originally designed for a Huret derailleur. Worked like a champ, and gave me a nice neutral position between 2nd and 3rd. :)

Speaking for myself here, but there is something about having a neutral in between gears that I would consider not working like a champ. I don't think that thing is working quiet right.:)

[QUOTE=carlton] I have shifted a Sturmey-Archer 3-speed with a friction-type downtube lever originally designed for a Huret derailleur. Worked like a champ, and gave me a nice neutral position between 2nd and 3rd. :)
We have had this discussion before. :) The problem is if you don't "center" the shifter properly you can quickly damage the hub. That's why geared hubs had "indexed" shifters way before derailleur gears did.

Done a lot of research on this and got a lot of comments on this forum. I concluded that Open pro/Ultegra hub which I use is way more serviceable and reliable for me than going the Internal hub route.

At the moment with the spares in my shed everything is serviceable.

I personally wish they would combine an internally geared and a dynamo hub into one unit.

I personally wish they would combine an internally geared and a dynamo hub into one unit.

Wouldn't it be better to spread the weight about a bit instead of having everything in the rear hub?

Wouldn't it be better to spread the weight about a bit instead of having everything in the rear hub?

The big advantage of a combo would be having only one wierdo hub needing a wheel built around it, instead of two.

Hmmm not sure I agree there - if you learn how to build wheels then switching hubs and rims only costs you as much as new spokes.

Argh!... I was just about to place an order for the Red Band Nexus 8-speed hub; now I need to reconsider because I'd hate for the SRAM hub to come out and make me regret the choice. Converting a bike to use the Nexus hub is an expensive proposition (the following are the cheapest prices I've found and assume that I build it myself):

Hub: $140
Rim: $30
Spokes: $25
Shifter: $25
Adapter to mount the shifter on drop bars: $55
(the shifter is only available in a version that mounts to a flat bar)


Total: $275!! Just looking at that price is already making me reconsider... ;)

Having one more gear isn't a huge deal to me (I generally ride fixed so 8 will already seem like a world of options), but I'm getting this to have the ability to quickly add gearing to my otherwise fixed gear commuter for light touring. Going from a gear range of 307% (the Nexus) to 340% (the SRAM) could be useful.

Ahh.. screw it. I've already mentally committed to the Nexus (and placed an order for the shifter last night) so I'm still going ahead. I'll write a report about it after I get significant experience with it.

-D

Hmmm not sure I agree there - if you learn how to build wheels then switching hubs and rims only costs you as much as new spokes.

Most cyclists will never build a wheel, and getting somebody else to do it for you is an added hassle and expense.

Whenever internally geared hubs come under discussion, out come the reduced efficiency statements.

I'm an evidence sort of guy. Is there a study that measured and compared said efficiencies that one can refer to?

And all these gazzilion gear ratios - how necessary are they? I've been commuting single speed and one thing that became clear is that once a suitable ratio is chosen, it works; what's more, on my derailer bikes I use less and less gears. I enjoy standing up to pedal on hills as I become more practiced to do so. Only when I do roadie stunts like TT do I use the whole cassette. So for that application, derailer for sure, but for all other cases? I wonder if we are being spoiled rotten.

I wonder if we are being spoiled rotten.

When I'm going up/down a few miles of hills having just the right gear is very important to prevent becoming prematurely fatigued. (i.e. in pain) I was a long time single speeder, and after switching to a 8/9-speed triple road bike now I've put behind me how I used to dread hills so much on the single speed, and how I used to have so much trouble understanding how people could cycle on them. If that's being "spoiled rotten", then call me a brat.

OK. You spoiled brat, you! ;)

I really like hill climbing, and the single speed has "given" me the ability to stand and pedal for many minutes at a stretch.

Argh!... I was just about to place an order for the Red Band Nexus 8-speed hub; now I need to reconsider because I'd hate for the SRAM hub to come out and make me regret the choice. Converting a bike to use the Nexus hub is an expensive proposition (the following are the cheapest prices I've found and assume that I build it myself):

Hub: $140
Rim: $30
Spokes: $25
Shifter: $25
Adapter to mount the shifter on drop bars: $55
(the shifter is only available in a version that mounts to a flat bar)


Total: $275!! Just looking at that price is already making me reconsider... ;)

Having one more gear isn't a huge deal to me (I generally ride fixed so 8 will already seem like a world of options), but I'm getting this to have the ability to quickly add gearing to my otherwise fixed gear commuter for light touring. Going from a gear range of 307% (the Nexus) to 340% (the SRAM) could be useful.

Ahh.. screw it. I've already mentally committed to the Nexus (and placed an order for the shifter last night) so I'm still going ahead. I'll write a report about it after I get significant experience with it.

-D

The 9 speed hub isn't coming until 2007 by the looks of the SRAM website. But the SRAM Spectro 7 is already better than the Shimano hubs when it comes to reliability, ease of wheel removal and smooth shifting. This hub is going to take SRAM well out in front of Shimano, but I'm sure it'll be quite a bit more expensive than the Shimano hub. The biggest improvement in the i-9 is that there's no longer an external click-box sticking out from the dropouts. The downside could be that the wheel will be a bit more ackward to remove.

Thanks for the comments. I did go ahead and commit to the Nexus "Red Band" Inter-8. I chose this over the SRAM based on positive comments from Sheldon Brown (the Red Band model is supposedly a big improvement over Shimano's previous offerings) and it is a little bit cheaper. After reading your comments I'm having second thoughts, but I'm going to go ahead and give it a try. My total cost (with shipping from three different vendors, and including a new tire and tube) is just shy of $310. Hopefully I'll have everyting in a week or so and won't have any trouble with the build. Wish me luck!

-D




The 9 speed hub isn't coming until 2007 by the looks of the SRAM website. But the SRAM Spectro 7 is already better than the Shimano hubs when it comes to reliability, ease of wheel removal and smooth shifting. This hub is going to take SRAM well out in front of Shimano, but I'm sure it'll be quite a bit more expensive than the Shimano hub. The biggest improvement in the i-9 is that there's no longer an external click-box sticking out from the dropouts. The downside could be that the wheel will be a bit more ackward to remove.

Wouldn't it be better to spread the weight about a bit instead of having everything in the rear hub?

Putting the Dynamo and internal gears on the same hub would probably not add that much weight. I would think putting weight on the drive wheel would be a net plus. I am not sure it is possible though, wraping wire around the internally geared hub is not a problem, but where to put the magnet?

I personally wish they would combine an internally geared and a dynamo hub into one unit.

Been done! Sturmey-Archer AG (3-speed) and FG (4-speed.)

Makes fixing a rear flat a real adventure...


Epicyclic gears are always going to suffer from excess weight, reduced efficiency, reduced reliability under heavy loads, and expensive repairs.
Derailer systmems are always going to suffer from inability to shift while stopped, reduced reliability in wet/dirty conditions, and a tendency of the chain to jump when the chain and rear sprockets get worn.

Sheldon "Epicyclic" Brown

+------------------------------------------------------------+
| How much do you hate riding that your ten mile ride |
| has to take you fifteen seconds less? – Grant Petersen |
| http://sheldonbrown.com/podcasts.html#grant |
+------------------------------------------------------------+

Hey Sheldon, When will the i9 be available?

OK. You spoiled brat, you! ;)

I really like hill climbing, and the single speed has "given" me the ability to stand and pedal for many minutes at a stretch.

I actually used to stand more than sit on mine - even on flat ground.

tendency of the chain to jump when the chain and rear sprockets get worn.


That's your bike telling you its long overdue for new parts. ;)

The 9 speed hub isn't coming until 2007 by the looks of the SRAM website. But the SRAM Spectro 7 is already better than the Shimano hubs when it comes to reliability, ease of wheel removal and smooth shifting.

Reliability? I haven't seen any evidence that the SRAM 7 is any more reliable than the Nexus 8.


Ease of wheel removal? Yes, the SRAM has a slight edge there.


Smooth shifting? No contest. No gear system I've ever used shifts as well as the Nexus 8. The SRAM is really not designed to be shifted under load.



This hub is going to take SRAM well out in front of Shimano, but I'm sure it'll be quite a bit more expensive than the Shimano hub. The biggest improvement in the i-9 is that there's no longer an external click-box sticking out from the dropouts.

Right, they've copied Shimano's superior system, placing the shift linkage safely inboard of the chainstay.

Whether it takes SRAM out in front of Shimano will likely depend on the ability or inability of the hub to shift under load. Shimano has a special invention that I don't entirely understand that helps disengage the pawls under load. This was first used in the 4-speed model, then the 8-speed (not the 7-speed.)

I'm looking forward to trying out the SRAM whenever it becomes available, but I haven't forgotten the disaster of their 12 speed Elan hub.

Don't count SRAM's chickens before they hatch...

Sheldon "Nexophile" Brown

+---------------------------------------------------------+
| Our country is the world - our countrymen are mankind |
| --William Lloyd Garrison |
+---------------------------------------------------------+

OK. You spoiled brat, you! ;)

I really like hill climbing, and the single speed has "given" me the ability to stand and pedal for many minutes at a stretch.

I particularly like having gears when riding in traffic, e.g. for fast starts at intersections. The "no shifting without moving" objection is easily circumvented -- if stopped in a high gear, simply lift the rear wheel, move a gear lever, and press down on the pedal. I normally don't use much more range than an English 3-speed's 50-90 gear-inches, but I certainly like having alot of steps within that range, with a granny gear in the low 40s in case I bonk or encounter a steep grade or a really strong headwind. I remain a huge fan of half-step and what Frank Berto calls "gear phreaking," with a 45-42 / 13-15-17-20-23-26 combo on my UO-8 commuter, which gives me 10 gears on top, a large-large cross-chain "no-no," and a very rarely used, but nice-to-know-it's there granny.

... Derailer systmems are always going to suffer from ... a tendency of the chain to jump when the chain and rear sprockets get worn.

This is no longer a problem for me, ever since some guy named Sheldon reduced chain wear detection to an exact science (the 1/2-percent solution). :)

Perhaps you missed the point here...?

Reliability? I haven't seen any evidence that the SRAM 7 is any more reliable than the Nexus 8.


Ease of wheel removal? Yes, the SRAM has a slight edge there.


Smooth shifting? No contest. No gear system I've ever used shifts as well as the Nexus 8. The SRAM is really not designed to be shifted under load.



I guess you have a lot of experience with these things where as I've only used the SRAM S7, but I'd based some of my comments on opinions expressed by some folks from cycling plus magazine here in the UK who have extensive experience with both SRAM and Shimano hubs. Certainly the S7 is much better than the 7-speed Shimano hub, so I guess Shimano must have made massive improvements in the new 8-speed hub, including extending the service life considerably. It seems the 7-speed Shimano hub needs stripped and regreased a little too often to stop the gears from slipping, but perhaps your experience is different?

Sheldon Brown Posts Earlier:

"Smooth shifting? No contest. No gear system I've ever used shifts as well as the Nexus 8. "


I was going to say I agreed 100%. Then I remembered that 4 speed auto nexus I test drove many years ago. I couldn't even feel that hub change gears. It would just become easier or harder to crank the pedals depending on whether it was shifting up or down. That was the only one I ever saw never heard anything else about them. Be nice to see a nexus 8 auto if the auto shifting part was dependable. Any way the Nexus 8 Certainly shifts smoother than my well maintained 3x9 Shimano LX front XT rear deraileur set-up. I hope the i9 is an improved design. A little competion among the manufactures results in better products for the consumers.

Whenever internally geared hubs come under discussion, out come the reduced efficiency statements.I'm an evidence sort of guy. Is there a study that measured and compared said efficiencies that one can refer to? .
I can't cite actual engineering papers but the reduced efficiency (mechanical loss) is well documented.

Derailer systmems are always going to suffer from inability to shift while stopped, reduced reliability in wet/dirty conditions, and a tendency of the chain to jump when the chain and rear sprockets get worn.

Sheldon "Epicyclic" Brown
Yes, and all the rest of the time they are lighter, more efficient and can cover a greater range of gear choices.

Commuters, beach cruisers, and transportation cyclists have enthusiastically rediscovered them in recent years, but I will be very surprised to see them used in racing or serious club riding, as they were in the 1930s and 1940s.

I agree completely. They are great for those applications, where clean, easy to use and hard to break are a bigger plus than weight, efficiency and harder-to-work-on-when-you-rarely-but-eventually-need-to are a minus. The inability to use quick release alone will keep them out of races, if nothing else does.

I have a cruiser with a nexus 3 speed coaster hub. It works well, but has always felt like it has a bunch of little bendy parts inside and sucks up power, as compared to the single speed coasters on my cruiser #2, 20" folder and road bike. Yes, this is probably my imagination.

I pulled the 3rd edition of "Bicycling Science" (Wilson, 2004, MIT Press) from my shelf. Check page 344. Some measurments are found there, cited from another work (Puckett, 1999) but gathered by the same guy under the same conditions.

In summary, for a clean new Nexus-7 hub gear transmission efficiency at 100W averages across the gears about 90%; high is 92% and low is 76% depending on the gear, somewhat randomly. Another measurement of that same hub had the low (in 3rd gear) at 86%, suggesting that this was an outlier. The overall average of the measurements taken by two different experimenters of the Nexus 7 hub seems generally that you're in the 85-92% range, depending on gear. If I had to pick just one number I'd say 88%.

For a clean new Deore LX triple (42-32-22) with a 28-11 cluster, efficency ranges from 99% on the low end (22x28)to 88% on the high end at 100W; the higher the gear the less efficient, generally speaking. For a gear of 32x18 or higher the efficiency was 94% or below, so for most riding I think that's the right number to focus on.

What's unclear is how much you're giving away with a gunked up chain in the two different cases, but I think that would be less favorable to the derailer system since the chain's got to make more sharp bends. What's more, I think it's clear that the Nexus 8 red band is a big improvement on the Nexus 7, so we can probably take the efficiency measurements of the 7 as a lower bound on what the 8 Red Band will offer.

I recently built up a Surly Karate Monkey with thie hub and I took it out this weekend on a 30 mile ride - up to then I was just tooling around the neighborhood on it. It sure didn't feel slow or inefficent to me, and shifted very nicely indeed. I would buy another one in an instant if only the shifter weren't so ugly looking and went on road bars more nicely - I think the right answer here is a bar-end type lever with a clear cable length adjuster gear indicator doodad partway along the cable. I think some dura-ace brifters have this (it's really quite inconspicuous) and the Nexus needs it because you need to know you're in 4th gear to properly trim the cable pull. I may buy another one anyway. It's a shame the Hubbub is so expensive.

To remove the rear wheel, you need a 15mm wrench to get the axle nuts off, and probably also a 2mm allen wrench to twist the shift ring around so you can pop the cable free. I went for the roller brake because I don't care about weight, so I need a 10mm wrench and a screwdriver to get that arm off. The cable for the brake comes off easily by hand.

However, you can get all these things in a multitool you'll probably want to be carrying around anyway (One of Park's has all that, although the 15mm wrench is crappy) and once you know how to get the wheel off it comes off without much fuss. The only thing is that if you don't know the right way to do it, the easy way isn't exactly self-explanatory.

I'm also inclined to think that the efficiency argument is a red herring. Yes, internal hubs will not be as efficient on average or at maximum as a derailer drivetrain system, but honestly, who cares? The difference is not significant enough for it to be noticed by your average rider doing the kind of riding that internal gear hubs are best suited for. Moreover, as bjkeen has said, a 'real world' derailer drivetrain is often dirty and gritty, especially if the bicycle is being put to the kind of use that internal hubs are good for - all-weather urban commuting and utility riding. The straight chainline and lack of derailer pulleys mean that a dirty drivetrain on an internal gear hub probably has a far smaller impact on the efficiency of the system. How bizarre that so many cyclists think that we should be concerned about such tiny efficiency gains or losses! The only people who have much reason to care about this are racers or recreational road riders (say that five times fast ;)). Internal gear hubs aren't too likely to make much impact here, but this is of no concern to 99.9999% of cyclists - why should we care about how useful something is to a racer? I'm not a racer - I'm a commuter, a utility cyclist, and an occasional tourist. Tourists are still stuck with derailers if they want an acceptable range of gears without dropping a 'G' on the Rohloff, but for my other needs, a Nexus-8 drivetrain would serve me better than my present derailer drivetrain.

Weight, too, is a non-issue. How much less does a derailer drivetrain weigh? Remember, an internal hub replaces a front derailer, rear derailer, freewheel/freehub cassette and one or two chainrings. The weight difference is minor, if it even exists. This is especially true considering that the kind of derailer drivetrain that this kind of hub will replace is NOT going to be the racy, top-of-the-line, 200g total weight deal that racers use. My drivetrain is a mix of mid-low level touring and mountain components. It has some serious mass! Most of the noticeable change is in the distribution of this weight: it will move further to the back of the bicycle. But even if there is a drastic weight difference, I offer the same objection as I do to the criticism about efficiency: who cares?. Ever do a search on "bike weight" in the commuter forum? It's safe to say that the average bicycle weight for BF commuters is hovering around 30 lbs, perhaps even on the high side of it. There are plenty of folks that break the 40 lb barrier, and even a couple of outliers up at the 50 lb mark. And remember that we're the performance-oriented, "enthusiast" commuters, hence our membership on a bicycle message board. If my bicycle will weigh a couple of pounds more when I swap out the current drivetrain for a Nexus-8 or i-9 drivetrain, so be it. Weight is one of the least important considerations for the utility-oriented cyclist. Low maintenance, reliability, consistent performance in all conditions - these concern me. A few hundred grams do not.

Rohloff is claiming 96%+ efficiency in all gears. If they can be believed, it's tempting... One also wonders how the efficiency will change over time, and the ultimate useful lifetime of a $1000 hub.

I'm also inclined to think that the efficiency argument is a red herring. Yes, internal hubs will not be as efficient on average or at maximum as a derailer drivetrain system, but honestly, who cares?

On the hills you'll care. Every little bit feels like the Demons of Gravity are hanging on your back wheel. Overall there seems to be enough momentum for internal gear hubs to eventually break through (again) in some non-performance class of bike.

bjkeen: Thanks for the info!
grolby: Nice arguments. Very valid. If internal hubs weren't so maligned by the majority (for no good solid reason) they would be more popular and cost less due to economies of scale.

The removal of wheel may be less of a problem than generally thought. As a boy I repaired my and my 4 sisters' bikes' punctures without ever removing the wheels. I just hauled out the tube after popping one tyre bead off the rim, pumped it in situ and found the leak. I did it the other day with my wife's bike (which has axle nuts on the rear wheel) on the trail and it was dead easy. Removing the wheel is far from necessary if you're prepared to wait for rubber cement to dry. While you are waiting, you find the puncture cause in the tyre, so hardly any extra time is spent versus replacing the tube.

Shimano has tried the internal gig for the masses and I'm sure they sell more of it in Europe and Asia than on this side of the oceans. It may be ok for the masses for commuting or pleasure riding. Due to a number of the things listed above and mostly weight I don't ever expect it to be widely accepted in the racing or road circles any time soon.

On the hills you'll care. Every little bit feels like the Demons of Gravity are hanging on your back wheel.
Well let's do some calcs to see how much you'll care.

Let's assume that you are expending 200W while climbing a hill. That is quite a good effort but within a regular commuter's capability.

Let's further assume that the efficiency of your derailer in that particular gear is 95% with somewhat dirty chain and imperfect chain line. The efficiency of the hub gear is 85% (a low estimate).

Using these figures, you are wasting 30W with the hub, and 10W with the derailer. These figures still are a bit abstract, so let's connect them to a slope and some mass.

From the Kreuzotter site, on a 5% slope, with a 15kg bike and 80kg rider, you are cruising at 11.7km/h with the derailer, and 10.5km/h with the hub gear.

On the flats, you would be cruising at 27.3km/h with the derailer, and 26.1km/h with the hub.

If you maintained 200W for the whole ride, and your ride was 30km far (say to work) of which 5km is the 5% slope, the total ride time with the derailer is 25/27.3+5/11.7=1h20min. With the hub, the time is 1h26min.

The difference in time is a mere 6 minutes.

What are you going to do with those 6 minutes??? You'll need more time than that cleaning and oiling the derailer chain.