Shimano Alfine 11 - Longer term reliability

[jb]

I'm inclined to think that a lot of the problems I experienced was due to grit / dirt getting into the pulley V. the timing on the pawls is very close and it does not take much to get them out of sync.
I started cleaning the pulley out before every ride and experienced a lot less problems.

IMO it would be better if the cassette pulley were of an open mesh design so grit could be pushed out.

The circlip holding the sun gear in plce is however an entirely different problem and is not likely to be experience if the bike is not subject to some serious hill climbing.

[Brucey]

I'm inclined to think that a lot of the problems I experienced was due to grit / dirt getting into the pulley V. the timing on the pawls is very close and it does not take much to get them out of sync.
I started cleaning the pulley out before every ride and experienced a lot less problems.

IMO it would be better if the cassette pulley were of an open mesh design so grit could be pushed out....

Oddly enough I have already made custom pulleys (mostly for weird derailleur mech modifications) from sheet metal that are 'open mesh' by design.

In essence the pulley is made from a flat disc; a series of short radial cuts is made at the periphery, (spaced at about 3mm intervals) and alternate 'petals' are bent either left- or right-wards. [if you are going fancy, a double bend in each petal can make for a better-shaped 'groove' and (if necessary) a reinforcing ring can be tack-welded (or brazed) each side. This isn't strictly necessary, but can make for a neater job, that is also less likely to injure you should you come into contact with it.]

The result is a 'pulley groove' that cannot easily clog with mud because it will always be pushed and then shed to one side or the other. I have not yet built such a pulley for fitment to a shimano IGH, but (barring the faff-munguous spline on the inside, and the cable anchor) I don't see why it couldn't be done.

It also occurs to me that a small plastic guard of some kind just above the pulley might help to stop the pulley region from seeing so much dirt etc anyway; on an offroader, as well as what is hurled off the front wheel, I think a fair amount just drops down from the seat stays, perhaps being knocked off the tyre at that point. As mentioned earlier (much earlier!) in this thread, even if it doesn't clog the groove, dirt causing the cassette joint to bind is commonplace anyway.

cheers

[SA_SA_SA]

.....
So from this, I'd provisionally conclude that the bad cable (and perhaps a bit of sticky shifter trouble too) had turned what ought to have been a pretty decent hub into little more than a source of trouble for its previous owner.....

Wouldn't putting the indexing in the hub avoid most of this trouble on many-speed hubs?
If it improved the perceived reliability of said hubs (when not maintained by a Brucey ) wouldn't that be worth something to the manufacturer?

[Brucey]

...Wouldn't putting the indexing in the hub avoid most of this trouble on many-speed hubs?...

maybe, (and in fact I agree it would have been worth trying) but it is as well to be aware of the law of unintended consequences....the hub internals are already very complicated, there's not much room inside many of these hubs for more 'stuff'.... also, I would say that if the cable control was bad enough, it could still lead to the mechanism sticking between gears, if the indexing is inside the hub.

A more trivial point is that many users like to have a gear indicator. For this to work properly would almost certainly require that the cable was in good condition, regardless of where the indexing is.

cheers

[ukdodger]

...Wouldn't putting the indexing in the hub avoid most of this trouble on many-speed hubs?...

maybe, (and in fact I agree it would have been worth trying) but it is as well to be aware of the law of unintended consequences....the hub internals are already very complicated, there's not much room inside many of these hubs for more 'stuff'.... also, I would say that if the cable control was bad enough, it could still lead to the mechanism sticking between gears, if the indexing is inside the hub.

A more trivial point is that many users like to have a gear indicator. For this to work properly would almost certainly require that the cable was in good condition, regardless of where the indexing is.

cheers

The Rohloff has indexing in the hub which works fine. Mine has never jumped in over 3.5k miles. Also if you adjust the cables right the indexing arrow on the shifter always lines up with the gear chosen.

[SA_SA_SA]

...Wouldn't putting the indexing in the hub avoid most of this trouble on many-speed hubs?...

maybe, (and in fact I agree it would have been worth trying) but it is as well to be aware of the law of unintended consequences....the hub internals are already very complicated, there's not much room inside many of these hubs for more 'stuff'.... also, I would say that if the cable control was bad enough, it could still lead to the mechanism sticking between gears, if the indexing is inside the hub.

My idea was that if the indexing is in the hub, the hub will only deign to change to a definite gear, no in between positions allowed... ie if the cable doesn't move enough it simply won't change rather than enter an intermediate position and destroy itself....

[jb]

Hub indexing requires the selector to be turned in both directions by the cable - so you need two of them ,as with the Rohlof.

This is problematic for neatness and selector design.

A cable can be made to push as well as pull but it needs to be totally enclosed at all points and may become necessarily heavy.

[barrym]

If Shimano recognised that cables for gear changing was a flawed design, maybe they'd build one driven by some sort of electrical system....hang on a minute;-)

[Winfried]

Hello

I've been riding an Alfine 11-based bike for a couple of years. It also has a Gates belt drive.

The other day, I removed the rear wheel to replace the bottom bracket. After putting everything back together, I noticed some cracking noises when pushing hard on the pedals, such as when starting from a full stop (I always backpedal when changing gears to go easy on the IGH). That's when I thought about checking the indexing, and lo and behold: The two yellow marks where significantly misaligned, and I had to loosen the barrel adjuster on the shifter a lot to get the two marks aligned again when in speed 6.

Then, out of curiosity, I shifted down to speed 11 to loosen the cable entirely, and noticed there's only 180mm between the center of the bolt and the very end of the ferrrule, although Shimano says it should be 184mm.

I have a couple of questions:

• How can indexing be affected so much after just removing and reinstalling the rear wheel?
• How can the Alfine 11 work OK even when only 180mm separate the bolt and the outter-most end of the ferrule?
• How can I check that indexing works OK in the different speeds as the cable + housing get older, ie. the cable no longer slides as easily as it used to?

Thank you.

[Brucey]

I think that the specified 184mm cable setting is a starting point; the idea is that you then back out the barrel adjuster on the shifter until the indexing is set correctly. The exact measurement will vary with a few things including the fit of the reaction arm onto the hub; a small amount of wear there can alter the cable length setting by 1mm quite easily.

Probably something (in the cable) moved and didn't settle in the same way again when you removed the rear wheel; such things happen often.

cheers

[Winfried]

Good to know.

What do you mean by "reaction arm"?

Is it this part in the cassette joint, or something within the hub itself?

Actually, the 184mm is not from the center of the bolt and the end of the ferrule:

Alfine11.Inner.cable.fixing.bolt.unit.jpg (7.39 KiB) Viewed 2950 times

I'll measure it again tomorrow to check.

[Brucey]

by 'reaction arm' I meant that part of the cassette joint that doesn't move.

cheers

[ukdodger]

Good to know.

What do you mean by "reaction arm"?

Is it this part in the cassette joint, or something within the hub itself?
Alfine11.cassette.joint.arm.jpg

Actually, the 184mm is not from the center of the bolt and the end of the ferrule:
Alfine11.Inner.cable.fixing.bolt.unit.jpg

I'll measure it again tomorrow to check.

Having owned one of these awful things the exact adjustment really depends on trial and error while riding. Turning the adjuster one or two revolutions in one direction and if it gets worse go one or two in the other. I did get it spot on once but swapping it regularly for a Rohloff to do a tour I got fed up with it. I know Brucey will disagree but I think there are just too many gears for a hub that size and adjustments are too critical.

[squeaker]

I know Brucey will disagree but I think there are just too many gears for a hub that size and adjustments are too critical.

IMHO it's the reliance on cable tension that's the issue - I'm sure a Rohloff type 11-spd hub would shift faultlessly

[ukdodger]

I know Brucey will disagree but I think there are just too many gears for a hub that size and adjustments are too critical.

IMHO it's the reliance on cable tension that's the issue - I'm sure a Rohloff type 11-spd hub would shift faultlessly

Yes it does.