Servicing DA 7900 hubs

[Ross K]

Have any of you good people stripped and regreased the bearings on your Dura Ace 7900 hubs?

The Shimano tech docs are not really descriptive enough as far as I can see. They are the handbuilt type, not factory hubs such as the 7900 C24 wheelset (although they may be the same, I don't know).

I'm aware that they are not the usual cone-spanner-adjusted type such as Ultegra and below, and I obviously want to get the assembly and adjustment right as they cost me an arm and a leg!

Anyway, any first hand advice would be much appreciated. Many thanks.

[Brucey]

see

http://bike.shimano.com/media/techdocs/content/cycle/EV/bikecomponents/FH/EV-FH-7900-2875A_v1_m56577569830728255.pdf

for the ev techdoc and

http://www.slowtwitch.com/Tech/Hub_How-To_-_Dura_Ace_7900_3601.html

for the procedure.

These hubs work just like any other cup-and-cone hub except that the balls are clipped and the cones are locked differently to conventional cup and cone hubs.

[Obviously the freehub comes to bits too, (and can be reshimmed if there is any play in the bearings) but those guys think 'a freehub service' is just squishing a bit of lube through it, and in truth that is usually enough if the bearings are OK.]


Annoyingly these hubs cannot easily be adjusted using a conventional hub vice (as described on Sheldon Brown) so whilst they don't need cone spanners I rate them as more fiddly to adjust than previous versions. IIRC you still need to set the bearings slightly slack when the hub is out of the bike, so that when the QR is tightened, the axle compression removes the last few microns of slack from the bearings. If you set them with no slack in, they tend to wear more quickly than they should, because the bearing preload is way too high once the QR is tightened.

hth

cheers

[Valbrona]

Can someone please indicate what size hex wrench is needed for the freehub fixing bolt on these particular hubs. Thanks.

[Ross K]

Brucey, many thanks, that's really helpful, that second link is very clear. I assume the C50 version has the same internals as the normal flanged bubs I have?

Are the front hubs the same, minus freehub of course?

Valbrona, according to the second link: "This requires a big-boy 14mm allen wrench"

[freeflow]

If these are the 'digital' hubs then they are a doddle. Be aware that unlike conventional cup and cone you don't need to have a slight amount of slack which is taken up by the clamping of the quick release.

[Brucey]

Brucey, many thanks, that's really helpful, that second link is very clear. I assume the C50 version has the same internals as the normal flanged bubs I have? Are the front hubs the same, minus freehub of course? Valbrona, according to the second link: "This requires a big-boy 14mm allen wrench"

if you look up the relevant ev techdocs you can see the construction of the hubs very clearly and you will find that many of the internal part #s are identical I would have said.

Yes the front hub is a similar construction.

BTW not much point in removing the freehub body IMHO; you can usually clean and re-lube it in situ well enough, (even if there is a rear seal). Anytime you disturb a freehub body there is a small chance that it'll work loose in service once refitted. I normally check the bolt is tight and leave it be; once all those interfaces have settled into one another nicely it is probably best to leave them alone if you have any choice in the matter.

If these are the 'digital' hubs then they are a doddle. Be aware that unlike conventional cup and cone you don't need to have a slight amount of slack which is taken up by the clamping of the quick release.

I don't believe this to be the case; the adjustment method of the cones cannot prevent the axle from being compressed by the QR load, can it?

cheers

[RRSODL]

IIRC you still need to set the bearings slightly slack when the hub is out of the bike, so that when the QR is tightened, the axle compression removes the last few microns of slack from the bearings. If you set them with no slack in, they tend to wear more quickly than they should, because the bearing preload is way too high once the QR is tightened.

hth

cheers

Contrary to Shimano guidelines, you do need to set the bearings slightly slack.

I don't find them any fiddler to adjust than the "non-digital" system.

What I would say thought is that you can easily over pre-load the bearings if you were to follow Shimano guidelines for adjusting the cones. So, in that sense, the digital system can give a false sense of security when it comes to adjusting them.

I adjust them using exactly the same method as if I were adjusting non-digital cup and cone hub bearings.

[freeflow]

It is supposedly a feature of the digital hubs that they don't need the slight amount of play. I now can't find the document that told me this. But in terms of the cone adjustment they are exceedingly simple requireing only fingers one you have released the cones with the 5 mm hex wrenches.

[Brucey]

It is supposedly a feature of the digital hubs that they don't need the slight amount of play. I now can't find the document that told me this.

there's also a lot of people who have moaned about premature bearing wear in these (and similar) hubs. Are these things related I wonder...?...

But in terms of the cone adjustment they are exceedingly simple requireing only fingers one you have released the cones with the 5 mm hex wrenches.

you can move the cones easily enough but this is only half of the 'adjustment procedure'; with these hubs you then can't tell if the bearing adjustment is correct or not until you have reassembled the QR and tightened it against something so that the axle is compressed. [A small amount of play should disappear and if there was none before you now have excess preload on the bearings, just as with any other cup and cone style QR hub.]

By contrast you can adjust a normal QR cup and cone hub using cone spanners in just a few seconds and know that you have done it right if you use a proper hub vice.

cheers

[RRSODL]

It is supposedly a feature of the digital hubs that they don't need the slight amount of play. I now can't find the document that told me this. But in terms of the cone adjustment they are exceedingly simple requireing only fingers one you have released the cones with the 5 mm hex wrenches.

You are right, I've read the same doxuments but as I said before, I agree with Brucey and it's necessary to leave a little play when adjusting the cones.

As for the simplicity of adjustment, requiring only fingers, you are right too, however, that doesn't mean that is all that is required to do the job correctly and that was my point on my previous post.

[Gattonero]

Once I saw a video of an empirical test, a hub with the bearings set right, then squeezed in a vice, it won't get any friction unless a lot of pressure (far more than what a QR skewer does) was done by the vice.

The compression generated by the skewer could increase the pressure on the bearings on old type hubs, with lower quality steel axles. Modern hubs have either ground steel or/and oversize axles, I cannot imagine a skewer being able to cause any compression in a ø10mm ground steel axle.

Yet, bearings should always be set right, better a little play than too tight, that is very true.
The new Shimano hubs are so easy to set, just needs two allen keys

[fastpedaller]

Once I saw a video of an empirical test, a hub with the bearings set right, then squeezed in a vice, it won't get any friction unless a lot of pressure (far more than what a QR skewer does) was done by the vice.

The compression generated by the skewer could increase the pressure on the bearings on old type hubs, with lower quality steel axles. Modern hubs have either ground steel or/and oversize axles, I cannot imagine a skewer being able to cause any compression in a ø10mm ground steel axle.

Yet, bearings should always be set right, better a little play than too tight, that is very true.
The new Shimano hubs are so easy to set, just needs two allen keys

So my old 70's Campag Record hubs must have been rubbish - Every day really is a school day

[Brucey]

Once I saw a video of an empirical test, a hub with the bearings set right, then squeezed in a vice, it won't get any friction unless a lot of pressure (far more than what a QR skewer does) was done by the vice.

The compression generated by the skewer could increase the pressure on the bearings on old type hubs, with lower quality steel axles. Modern hubs have either ground steel or/and oversize axles, I cannot imagine a skewer being able to cause any compression in a ø10mm ground steel axle....

well you probably need a better imagination then.... .

The 'low quality steel axles' in fact weigh about the same (or are slightly heavier than) the 'oversized aluminium axles'. Given that the specific modulus of aluminium is almost identical to that of steel, both axle types are liable to deform elastically to a similar extent under the QR pressure. If there is any weight saving in the aluminium axle, it will definitely deform more than the equivalent steel part. [Shimano may be touting this hub system as an improvement but I think it might actually be worse in terms of unintended bearing preload.]

How much load is on the bearings because of QR pressure? Well (given a few givens) if you start with zero clearance in the bearings it is likely to be in the region of 100 to 300 kg for a moderately tight QR skewer, i.e. considerably more than the service loads that would normally arise in the bearing (the exact loads will vary with the stiffness of the hub vs the axle as well as how tight the QR is). If the bearings are good quality you won't see 'extra friction' because the CRR of a new bearing is very low and any crappy 'test' you might do at low speeds is liable to be dominated by seal drag anyway.

That you don't see 'extra friction' doesn't mean that the bearings aren't knackering themselves though; such loads are almost certainly well in excess of those required to cause premature fatigue damage in the bearing races.

To replicate those loads in correctly adjusted hub bearings, you would need to ride your bike around with about five fat blokes sat on it. You don't need to be a towering genius to work out that the bearings won't be designed for that kind of service....

cheers

[RRSODL]

The new Shimano hubs are so easy to set, just needs two allen keys

It seems I'm hitting my head against a brick wall......

Last attempt.....

Adjusting the new Shimano hubs correctly is not different to the old style, besides the need for cone spanners

To check your bearing adjustment, put the wheel into the bike with the quick release just barely tight enough to keep the wheel from falling out. If you are working on the rear wheel, take it out and re-install it so that the chain is not engaging the sprockets.

Try to wiggle the rim back-and-forth between the brake shoes; since the QR isn't tight, there should be a bit of play. If there is, hold the tire so that the valve is at the 3:00 or 9:00 position, then let go of it. On most wheels, the valve is the heaviest part; on wheels that have spoke reflectors, the reflector will be the heaviest part. Whatever is the heaviest part of the wheel, it should cause the wheel to turn and swing back and orth like a pendulum, before finally coming to a stop.

Once you have seen how the wheel turns with the quick release loose, try tightening the QR, then check again. If your bearing adjustment is correct, the play will disappear, but the wheel will turn as freely as it did when it was too loose. For very fine tuning of this, you can slightly vary the adjustment of the quick-release skewer, as long as it is good and snug.

[Brucey]

It seems I'm hitting my head against a brick wall......

Last attempt.....

I feel the same way.

As you, Sheldon, (and anyone else I have any time for as a bike mechanic) all say, if you don't leave a little slack that just disappears when the QR is tightened with these hubs (or indeed any QR cup and cone hub), you will be knackering them prematurely....



cheers