I'd like to see the scientific inquiry showing that derailleur systems are less efficient if you can point me in that direction. Like I said before, everything I've seen to this point says the opposite. I'm looking for something that talks about more than just the pros and cons of various cog sizes.


My commute is 75% flat, especially the part that's in traffic. I spend the majority of time in the middle range of my gears. I can usually start pretty easily in any gear I happen to be in even if it's not the ideal one. Not always, but most of the time. Being able to quickly transition from lowest to the highest gear has almost no value since I would rarely choose either one for any part of my commute.

Interestingly enough this advice is the opposite of what is practiced by professional mechanical engineers who purposely chose odd toothed sprockets in order to distribute the wear more evenly:
http://www.renold.com/Products/Trans.../Sprockets.asp

I've never tried Sheldon's recommendation before, but as we speak I am putting together a hub gear bike with an even rear cog (largest available) and an even chainring (on sale). I'm curious as to which advice is better (Sheldon or the engineers), but it wouldn't surprise me in the slightest if it were the less credentialed of the two. Either way I expect the chain will outlive my patience to carefully test it.

Cheers. :thumb:

Those same mechanical engineers you like to reference also have recommendations on chain lubrication requirements and when it comes to things like this, Sheldon Brown's credentials were rather excellent.

It would be interesting to test this to see just how many more miles one could get out of a drive set up like this... my winter bike runs a 40:20 and after an entire winter has no measurable chain wear. I am using a KMC 710 1/8 and have found these run for an extremely long time and only start to show minimal wear at 6000 km after some pretty extreme use.

Might have to mark the sprocket the next time I pull off the wheel and figure this chain will take me through summer rains and well into next winter... this is a plus with an IGH and a good chain / sprocket as the chain life is far better than any derailleur equipped bike that has to handle the same conditions.

This study reports a 5% increase in efficiency for bigger cogs and a 2% decrease from standard cluster to IGH with small cogs:
http://www.ihpva.org/HParchive/PDF/hp52-2001.pdf
So if you combine the two the net increase in efficiency is ~3% for a properly configured hub gear (maybe not quite that much because you can't get IGH cogs that big...I estimate it's closer to ~2% for stock cogs).

Other studies have reported similar results.

You might feel differently if it were easier to shift between widely dispersed gears (it doesn't necessarily have to be THE lowest and THE highest gear...just as easy to use 2 and 7 or 3 and 6 or whatever tickles your fancy...once you get good with the shifter you just slap it and it magically goes to the gear you want). Personally if I were going to be starting in less than ideal gears I'd sooner take a singlespeed bike than a derailleur (why bother having them if you can't use them to the fullest?). To each his own.

But I agree that it's not that hard to get into the right gear with a derailleur before stopping...it's just that I'd rather be focusing on braking, filtering, whether I can beat the yellow, picking my nose, etc. I also often change my mind about what gear I'm going to start with once I'm done picking my nose and have time to inspect the condition of the intersection (amount of potholes, etc) or if I decide to take a shortcut (right on red, maybe move to left lane for an upcoming turn...or maybe even run the light if no cross traffic).

Yeah, well there's no substitute for trying it yourself. Sheldon knew that very well: Some of the stuff he did should make anyone cringe, but there's no arguing with success.

As to the lubrication (if we'll ever let it die), some of the engineering requirements also relate to cleanliness. If we listened to them we'd have to conclude that roller chain is simply unsuitable for exposed bicycle use.

Let us know. The one I'm building is 44/22 (not enough clearance for 46/23 or 48/24), but this one's a recumbent so the carbon frame will probably delaminate before the chain even begins to show signs of wear.

You mean the races where the racers need to be able to pedal full out, without pause, and still shift through several gears while doing so - something an IGH is not as good as a derailler at...

With a derailler you use a technique called "downshifting a gear or two as you come to a stop".

Only against people who don't downshift when they're coming to a stop. For most of us we would smoke you as we don't need to wait, or put a pause in our pedalling in order to upshift away from the light.

Not a big enough dead spot for the igh to always immediately shift while under load.

There are other points that I didn't quote, I'm not ambitious enough to go look them up.

If you have hung around here around here long enough you would see that when it comes to setting up singular drives I always advise using the largest chain rings and cogs that are practical as it has been known for a long time that smaller cogs decrease efficiency and drive train efficiency drops off markedly when you use a cog smaller than 13 teeth.

A singular drive with a clean chain and proper chain alignment will lose about 1% and this is what makes a fixed gear bike the most efficient bicycle from a mechanical standpoint but not as efficient from a practical standpoint as humans have such a narrow power band.

We have to include the engine in all of this...

Across a multiple speed cassette the frictional losses will vary widely and you may only lose 1% when you are running in those positions where your chain is straight and cog sizes are reasonable and will lose more when you are using the cogs with less than 13 teeth.

Losses approaching 20 % can be observed when you cross chain the drive depending on which way you are cross chaining.

If used correctly a derailleur equipped bike can also be extremely efficient as it allows riders to match their output to the conditions and a wider gear range more than offsets minor efficiency losses and if you can shift effectively efficiency loss can be kept to a minimum.

With an IGH you run the same straight chainline and because the drive gears are sealed the mechanical losses should remain consistent within the hub and are quite low... with any hub gear the direct drive is the most efficient as there are no moving parts and it runs like the fixed gear bicycle. Move away from the direct drive position and the efficiency will drop a little.

Modern hub gears are getting better and better and if the Shimano 11 speed proves to be successful it will come very close to offering the range and smaller steps you get out of a derailleur equipped bicycle at an affordable price and he price of the Nuvinci has come down a fair bit.

Most people's complaint about 8 speeds is that the steps between gears are too large and that you sacrifice top end for a lower gear or bottom end for a better top gear... this also is a factor in overall efficiency.

But doesn't it say that "Hub gears are generally about 2% lower in efficiency than derailleur-type gears" ?

(Item #2 under conclusions).

That's pretty close to the number I've heard before.

I think this talk of efficiency is rather vague and largely unimportant to the average commuter, especially when it's on the order of a few percent.

As far as I can see, both have practical advantages:

IGH:
-chains can last longer, since full-bushing chain can be used with a straight chainline, and can be fully enclosed in a chain case.
-use of said chain case prevents grease marks on clothing, making the bicycle more suitable for commuting in everyday clothes.
-can shift while stationary.
-simple operation due to only one shifter.
-generally less demanding of maintenance.
-rear wheel requires minimal or no dish.

Derailleurs:
-potential for a wider range, and smaller steps within that range
-can shift under load with more modern systems.
-can customise gearing to a large extent, unlike IGH, where the steps are fixed and only the 'final drive' ratio can be changed.
-simpler to repair at home.

I don't mind IGH steps when they're uniform and smaller than 20%. This is what makes my SRAM i-Motion 9 my favorite road IGH. 17% steps, 340% range for ~$350 with shifter. It's not as tough as a Shimano 8 speed but it's good enough for the road. Shimano 8 speeds have a 25% step between 5th and 6th gear. That translates to going from 95rpm to 70 rpm when I shift from 5th to 6th, I find it very irritating on the road. I'm not comfortable with a cadence less than 80rpm or more than 95rpm when I'm pushing, it puts a hole in my pace speed.

This my biggest gripe about the Alfine 8, as I spend the majority of my time in 6th gear, and use 5th for smaller grades or brisk headwinds (4th for ripping headwinds). This is a design oversight that it appears the 11 speed has eliminated, with its dead even spacing. I'm going to wait it out a year or so to let Shimano work out the bugs before I move towards a wholesale replacement of my current Alfine 8 setup.

That study is not high end, it has the worst of engineering reductionism/determinism (text like "For some reason that is not apparent," is a sign that their underlying hypothesis of what the heck is happening has gaps) and no accommodation for uncertainty in measurement (likelihood of greater variation between trials than between treatments)

The Paragon tube splitters are offered in SS or Ti. As the SS ones cant be brazed with brass rod (per Paragon) how do I join these to my steel frame? Silver solder? TIG? Chewing gum?

I suspect the needle and ball bearing lavished Rohloff , and a plain bushing internal
like a Sturmey are different , but the trace differences in efficiency
is a merit less comparison in the real world .

It's transportation after all , if it gets you there a half second slower
due to an .05% increase in some bearing surface resistance parameters,
then so be it, relatively , worse if you don't catch all the green lights
on your route.

Disclaimer: I'm a derailer(RIP Sheldon) guy. This due to never riding an IGH, though I'd love to ride a "nice" one and perhaps I'd become an IGH guy, though I don't see a need.


I've never had a problem with the derailer system. Once set up, It does it's job and is easily maintained by myself. It's a very simple mechanism that sheds outside contaminants pretty easily since it's all exposed. In the wet they work flawlessly. Those crazy MTB cats put derailers through all kinds on hell and they stand up to it. Go home, hose them off and re-lube and you're ready for the next time. A steel cassette lasts plenty long, and steel rings are available and last a very long time as well. Derailers don't have to be expensive to work. Even the cheapy cheap ones do their job. If gumming up of shifting is a concern, then friction shifters that work regardless of muck and grime can be had for wicked cheap.

I'm also curious about the affect of riding a longtail bike(Xtracycle, Big Dummy, Yuba, Madsen, Kona Ute, etc.) on derailer efficiency because of the lesser angle of engagement from chain to cog/chainring due to further distance between them. (I ride an xtracycle-equipped MTB for daily rider, as well as a Cross Check and the Xtra seems less affected by cross gearing.)

When I was looking for a winter bike I wanted a chaincase - seemed really cool, would definitely keep all that road grit off the chain, and I can definitely see stuff lasting longer with one, you know? :-)

The problem, sadly (in the US at least), is finding one. They seemed really cool, but the only bikes that seemed to come with them were either the more expensive Breezer models (which I could not find in any store anywhere in the Minneapolis/Minnesota area and I wasn't going to buy a bike sight unseen), or really cheap "cruiser" style bikes meant for only the most casual riding. :-(

Always seemed like it would be really cool to have one though...

Unless I'm misunderstanding what you're saying, if you're spending most of your time in 6th gear you could fix at least part of your problem by putting a larger chainring on the back. If I remember right, 4th gear is the direct drive gear, if you put a bigger cog on you could spend most of your time in the 4th gear and still have plenty of range on both sides?

5th is Direct Drive, and I totally agree that a slightly smaller cog in back would allow me to stay in 5th rather than 6th.
This is one of the inherent limitations at the moment with the belt drive system, as the only rear cog that I can get for the Alfine is a 24T. I do have options up front (I am currently using a 50T and have a 55T in my spare parts bin) but I actually like my current gear range (29" low gear, 90-something high gear) for the area I live in (Rocky Mountains).

I'll stick with it until I feel like the 11sp is dialed in (or just bite the bullet and get a Rolhoff).

Despite the relatively new application the belt drive represents (to the bicycle world anyway), I have truly enjoyed the system on my bike and feel the combination of a belt drive and an IGH is a fantastic match. I know (as with any new application) the system will go through its own evolution, and I look forward to what the future brings.

Chucky - Based on your comment below, I guess I am a fool :D So be it.

I was looking forward to this new 11 speed... as a bike builder and rider of 20 inch wheels it offers pretty much everything a touring cyclist or folder owner could ever want and the more even steps solve the issues that the 8 speed has.

It just has to prove itself as being bombproof and my experience with other Shimano IGH drives has not encouraged me... I see too many of them in my shop that need overhauling and replacing the internals is a $140.00 touch.

Am going to go and build a new wheel set for my folding touring bike and will be using an 8 speed cassette with the existing triple... will give me all the range I need and because being able to service the bike in the road the derailleur system makes great sense as none of the parts are hard to source. I could run a double but the triple gives me the cleanest and straightest chain lines.

Know from experience riding this bike over many thousands and thousands of miles that my 7-8 hour days in the saddle would not be as pleasant with an 8 speed IGH as the small gear steps I get with a close range block really help with stamina over 110-160km a day.

I usually tour at much higher speeds and rpm than most... my friends consider 20kmh a nice pace and I am often dialing it up to the high 20's and and can maintain this all day.

With that... I am off to ride my folder as it is actually too nice to be sitting and discussing cycling when you can be cycling.

I don't know Paragon's specs, but stainless is usually brazed with a high-silver filler, often with nickel, e.g. 50N -- 50% Ag/2% Ni, or BAg-7, 56% Ag.

Stainless is harder to braze than other steel alloys and has to be silver brazed with rod that is silver with added copper and tin with no cadmium.

It is one of the most difficult materials to work with as it requires careful preparation and exact heat control to prevent anealing and brittleness which will reduce the corrosion resistance and may cause failures.

The optimal method to work stainless is to TIG weld it as the gas shield prevents oxidization and the heat control can be very precise.

Yes I already said that. Add it to the 3-5% increased efficiency from setting it up properly for the final result....assuming you're capable of setting it up properly.

Right...so where is the study you published on drivetrain efficiency? [Redacted immature insult, I don't expect to have to do this again. Capeche? Walter, Moderator]
The results of the study I posted are not unique. It just happened to be the most easily accessible link.

You've done a poor job at extrapolating those results to the real world.

I set the range (30" to 92"), then I pick the primary gear ratio (38x18). The 5-6 step is unavoidable in any reasonable use case....

22T rear cog is available http://www.vaust.com/catalog/gates-c...b73vmisrpu9as6
I am currently using a 50T/22T in my commuter.

Good to know. I'm certain that more options will become available with time.
I accepted the inherent limitations that come with any new application in order to try it for myself. The only thing keeping me from slapping the 22 on there now is the knowledge that I'll need to do battle with the frigging lockring ;)

What's insulting is HardyWeinberg's attitude towards the hard work of those who have actually performed experiments and published their results. Not that I expect you to care because, after all, the authors of the study aren't part of the herd your advertisers are paying you to keep blithely ready to consume.

:lol: Because you say so? The proof of the pudding is in the eating. The real world speaks for itself.

No, "HardyWeinberg" is correct. You may not understand his arguments against the quality of the study, but they do nail some of the more general problems with this study. Fx. the lowest efficiency measured on the triple crank is in the middle ring with the straightest chainline. That result is hard to explain, so they don't, and that is a major problem, because this error propagates through the entire study. Taken at face value, this study shows that deviating chainlines (cross chaining) plays little if any role in efficiency (probably true) but that the straightest chainline actually lowers the efficiency (probably untrue). In any case, this clearly systematic error makes the entire study very problematic to use.

Besides that there are numerous small errors and inaccuracies like "Ultegra Mtn. Grupo" (Table 2. p.11): There has never been a Shimano Ultegra MTB group, and "grupo" is neither English (group) nor Italian (gruppo).

As a more general note, I am personally always vary of studies funded by commercial companies like the above study (Browning Research made bicycle gears). Especially if the one of the authors, like in this case, previously have written very favorably about the funding company's products.
Such studies tend to be both biased and made with crass commercial agendas.

Take the "research studies" showing that the breakfast meal is the most important meal of the day so one should always eat a hearty breakfast. Take other studies that clearly contradicts this. Now guess which studies are funded by major breakfast cereal companies and which are independently funded.

--
Regards

Um, no. More likely it is due to a lack of logic and reasoning skills. I'm just pointing it out.

For a couple years, I did a 38 mi r/t commute on a Nexus 8 hubbed bike with all the commuter geek stuff, like VO fenders, rear rack, dynamo hub, etc. I liked I a lot. Then I built up a derailleur equipped "fast" commuter which is stripped down to just a handlebar bag. Last year, it saw the bulk of commuting duty--bigger gear range more appropriate to my hilly then flat commute, closer spaced gears.

While riding that bike, I got a different frame for the IGH ubercommuter, and set it up with a tensioner and two chainrings up front. Sold it before putting any miles on it to finance frame building class.

Now building my next Ultimate Commuter, and it will be a derailleur bike.

If I was doing a city commute, I'd probably opt for IGH. There will be an IGH bike in my stable at most times--currently working on an S2C r20 project; I'm eyeballing some pretty sick closeout prices on Gary Fisher Simple City bikes and last year's color Belleville--but these are mainly for-fun or around town bikes.

For my longish, rural commute, I'm going with derailleurs.

As a mechanic who has worked on and ridden both systems, one is better than the other only in context of intended use, or more accurately, personal preference for intended use; they both have their advantages and disadvantages.

All-state 11 brazing rod with the right flux for stainless should do it. The local welding shop here said a pound of the nickel rod would cost me 30$. ESAB is the company that makes all-state rods and they make a flux for stainless too... call a local welding shop and talk to them. I've yet to try this myself so I am not vouching for it per se, but there are other options besides expensive silver rod.