Cycling UK Forum

Hi all

I have a Marin Point Ryes and just checked the bearing in the head set (for the first time in tried this operation ! ) and they were in a cage which was all rusty and most of the bearing dropped out.

Took to my local bike shop who could really match them to anything in stock so I ended up buying some loose bearing.

Question is can I easily replace the loose bearing case with sealed bearing ?

It would seem replacing would be easier and thought it maybe the way forward looking at most clips on YouTube !

Would appreciate any advice....

Many thanks for your help and time

Regards

Mike

If you have a bag of loose balls the same size as the balls that fell out of the "retainers" , then just clean it all up and stick the new balls in place with grease, and you are good to go.

loose balls all the way... just be sure to get the right number in there. One too few is OK but one too many is very bad. You will normally get 10-25% more balls in than the clip originally held BTW.

cheers

Why is it that the number of balls in a race should always be an uneven number?

pioneer wrote:Why is it that the number of balls in a race should always be an uneven number?

if the races are perfect then it isn't really necessary, I think, not at low speeds anyway. IIRC there are quite a few cartridge bearings designs which have 8 balls in them.

[edit; like this

In deep groove cartridge bearings the number of balls is dependant on the loading requirement and the design of the bearing, it seems.]

If you have imperfect (bicycle) bearings running at low speeds then I can see the logic of an odd number of balls; you will never get two balls diametrically opposite one another which might cause a really tight spot.

An odd number of balls may always support the load in a more stable fashion, where an even number may not, in the same kind of way as a three legged stool will always be steady on an uneven floor, but a four-legged one won't.

cheers

pioneer wrote:Why is it that the number of balls in a race should always be an uneven number?

It isn't.

They don't. Not always anyway. For whilst there may seem to be more odd-numbered examples on a bicycle, front hubs generally had/have 10× 3/16 balls per side and I've found even numbers in many headsets, freewheels and pedals.

Thanks, it's just that I recall changing BB bearings years ago and it always seemed that there was an odd number in the races.

I may be wrong, but aren't the cups different for sealed bearings compared with open/caged bearing cups?

Or is it a case of cartridge bearings not being picky about cup surface?

no cups at all in most cartridge bearing hubs...

cheers

Well that's kinda what I meant, in that cups for loose/caged balls aren't flat are they? so is dropping in a cartridge bearing even an option?

not an option; not without a lot of work.

A standard hub with a cup and cone bearing has a cup race pressed into a recess. The cup may be curved on the back, but the recess is usually square.

if you can get the cup out you will be left with something that looks like a hubshell designed to accept a cartridge bearing. However there are three snags;

1) the cups don't come out easy. Sometimes you can drift them out but more often the quickest way is to weld something onto the cup in situ and then beat the bejeezus out of it.

2) you will need a different axle. Axles for cartridge bearings are usually quite different from those meant for cup and cone bearings. You need to have a machine shop handy or to be able to buy new parts. New parts that fit; these are not commonplace.

3) The bearing recess may not be the right size for any cartridge bearing. Yup, hubs are made many different sizes, most of them weird. If you get lucky and your hub has a 30mm recess (there are a few that are made this way) then there is still no guarantee that the recess will be the right tolerance to accept a 30mm cartridge bearing. In fact it is more or less certain that it won't be good enough; unlike a cup insert, the fit for a cartridge bearing needs to be right to microns.

So I have remachined hubshells, machined up bearing carrier adaptor inserts, and even ground bearings down so that I could fit them into hubs that were never meant to accept them, made up axles and so forth. I have made some nice hubs this way, in most cases they are ones I couldn't have bought; if I could have bought something similar I am absolutely certain that it would have been a darned sight quicker to have just bought another hub like that and rebuilt the wheel etc.

Like I said at the top of this post, it is a lot of work.

cheers

pioneer wrote:Thanks, it's just that I recall changing BB bearings years ago and it always seemed that there was an odd number in the races.

Yup, cup and cone bottom-brackets are almost invariably 11× ¼in balls per side. Except when the balls are in a cage, then the fingers of the cage take up some of the space, reducing the number to as few as 7 in some cases. With fewer balls to carry the load, the stress on each individual contact point is increased and with it the number of revs to failure reduced. With only 7 of 11, bearing life will be greatly reduced, so those caged sets are best thrown away and the full complement of loose balls used instead.

Please note the terminology. A set of balls, held in a cage, is not a race. A 'race' is the surface on which the rolling elements of a bearing roll, in this case the outer race is a cup and the inner race a cone.

There are three reasons for caging the balls in a bearing:
1) because it's impossible to assemble that type of bearing with a full complement so a cage is needed to space them out evenly
2) to halve the rubbing speed and insert a lower friction material between adjacent balls, so as to avoid overheating and consequent lubrication failure at high rotational speeds
3) for convenience in assembly.

1) The first reason does not apply to cup and cone bearings: it's always possible to assemble these with a full complement. (For an example of a bearing this DOES apply to, see the deep-groove ball bearing illustrated by Brucey with 8 balls in. The only way to assemble such a bearing is with the inner race off-centre, half fill with balls, re-centre the inner race, spread out the balls then add a cage to keep 'em like that.)
2) Everything on a bike (with the possible exception of a bottle dynamo) turns very slowly in engineering terms, so ball-on-ball friction is insignificant.
3) So the ONLY reason for caged bearings on a bike is assembly convenience. And I don't think that's a good enough reason to let things wear out sooner. I reckon I'll save more of my time by not having to mend the darned thing again, than I spend by poking balls one-by-one into the grease. So I always ditch the cage and fit a full complement.

You can always have your cake and eat it and use loose and a ring of balls.
Had someone yesterday asking do adjust his headset (old cheap to middling Raleigh ATB). Wouldn't adjust (stiff and loose at the same time), I noticed I could just see the bearing cage (ring of otherwise loose balls) at the bottom cup. Lowered the fork a bit (unscrewed lock nut/cup a bit), only to see another set of loose balls above the bearing ring, top race was the same.....
I guess even with balls, you can have too much of a good thing!

You can always have your cake and eat it and use loose and a ring of balls.
Had someone yesterday asking do adjust his headset (old cheap to middling Raleigh ATB). Wouldn't adjust (stiff and loose at the same time), I noticed I could just see the bearing cage (ring of otherwise loose balls) at the bottom cup. Lowered the fork a bit (unscrewed lock nut/cup a bit), only to see another set of loose balls above the bearing ring, top race was the same.....
I guess even with balls, you can have too much of a good thing!