Thorn Sherpa, Co-Motion Custom Road, Salsa Fargo, Mercian King of Mercia and Motobecane Fantom Pro 29er
I am hoping to get the definitive answer hear as I have been told different things by different shops. Hope this isn't a dumb question.
I have been told:
A) When adjusting the cones, allow a little play so that when you clamp down on the skewer, it will bring the cones together and eliminate the play. If you don't, the cones will be too tight and will cause problems. To do it right, adjust the cones, clamp on the frame and check for looseness and adjust as needed.
B) Don't do the above for high quality hubs. Adjust the cones so there isn't any play because when you clamp down on the skewer, the cones aren't going to budge.
'''96 Litespeed Catalyst, '05 Litespeed Firenze, '06 Litespeed Tuscany, '12 Surly Pacer, All are 3x8,9 or 10. It is hilly around here!
Leave a tiny (emphasis: TINY) amount of play in the hub when you assemble it then check to be sure there is no play at the rim after it's clamped into the dropouts.
High quality hubs differ from low quality hubs in that their steel axles and cones are harder, stronger and better finished but their elastic modulus (the amount they compress under a given load) is identical to the cheap ones.
Another issue with newer hubs is the lightweight skewers - they stretch rather than compressing the axle slightly as with older designs. I was having trouble keeping a new wheel aligned in my chrome hardened horizontal dropouts on one bike, and my LBS told me to find an older (heavier) skewer. Problem solved with an 80's Shimano 600.
For this reason, too, you don't leave as much play.
I vote for A. Leave a tiny amount of play. How much? You'll have to experiment and determine by trial and error. The trial and error process is a little bit of a pain, but worth it - this is a sign of a good craftsman. With the wheel mounted, you can easily feel for play, since the distance from the cones to the wheel accenutate the play.
(see http://wildavis.smugmug.com/Bikes) Bianchi Veloce (2005), Nishiki Cascade (1992), Schwinn Super Sport (1983)
This link will explain all you need to know (scroll down to "hub-adjustment"). My method differs from theirs in that I use a 5/6" open wrench on which to clamp the QR instead of the rear fork, but apart from that the idea is exactly the same. It's rather fiddly, but it's worth doing.
It's all trial and error - different things work with different setups. My advice: leave a little play, clamp the wheel into your dropouts as you normally would, and see if there's any lateral movement at the rim when you push. If so, readjust. I don't think there's any one theory that works all the time; you just have to know what works for your set of wheels.
Yeah, pushing on the rim sideways between the brake-pads will let you feel how much play there is. Quite obvious when you have the QR barely on vs. when it's fully tightened down.
I vote for option-A. Regardless of the "quality" of a hub, the axle is what gets compressed by the skewer and the cones on each end will end up moving closer together. Compressibility and Young's modulus of all steel alloys are pretty much identical. What you will notice is that some hubs have tigher-fitting cones to the axle and will spin on easier with less wiggle. However, once the locknut is tightened down, the entire axle assembly is solid.
The actual amount of play to leave is then trial & error. You get a good feel for it over time. I've found that it's usually about 20-30 degrees of a turn looser from the adjustment where there's no play.
Sheldon Brown also has a great article on cone-adjustment. He made a special QR that will compress the axle without dropouts or washers needed. Allows for precise cone-adjustment without too much trial & error.
Originally Posted by aroundoz
I have been told:
Try this out yourself. As AndrewP said, you can use a couple of washers to simulate the thickness of the dropout (about 6mm for front) and then you can clamp the QR on the axle off the bike to make adjustments :
With no QR clamped, there's radial and axial free-play :
WIth the QR installed but not clamped, there's still play:
With the QR clamped halfway, you can feel that the play is still there, but reduced significantly:
Finally, with the QR fully clamped down, there's no play in any direction, yet the axle still spins smoothly: