DMF 

I can see that in a shop setting you might choose loose balls simply because it's a PITA to have to stock all the different carrier variations.

well biked

It does seem to be a very debatable subject at least. I contend that to get to the point that you have any real performance benefit from a retainer, you need a certain type of retainer, the cheap tin retainers usually seen on bikes just ain't gonna do it. That's my primary contention, and I think I'm in the majority in my thinking on that, subjective or not.

And I still say a very good question is this: why on earth would any manufacturer have ever used loose balls in hubs if their engineers didn't feel like there was a performance benefit over caged balls? Headsets and bottom brackets, very tedious and somewhat more time consuming to pack with loose balls than with cages, so that, to me, explains those applications. But why, except on these more recent Dura Ace and Campy hubs with the different type of retainers, have quality hubs always been packed with loose balls from the factory?

DMF 

Anyone know a bearing engineer?

There are other factors besides the obvious. In my discipline we call them "non-functional requirements". Ease of manufacturing would be a non-functional requirement. So would marketing requirements, like the resistance of mechanics to captured bearings.

well biked

And what do you call "guys who read something on 'wiki' and then decide the tried and true version of something is wrong," in your discipline?

But seriously, have you ever even seen a hub with retainers in it, in person? I haven't. And have you obtained the right sized retainers and put them in your cup and cone hubs?

cny-bikeman

Here's one for the books. In my wrenching days I more than once came upon a cheap hub with retainers and a dust cap integrated into the hub. had to cut the retainer into pieces to OH the hub.

DMF 

well_biked, I'm not advocating captured bearings (the engineers and Campy and Shimano seem to be doing that well enough). I'm exploring the "common wisdom". It may have had a valid foundation when it became commonly recognized as wise. It may still have a foundation. But if we can't remember what it is, we don't know whether it's still valid (assuming it really ever was).

There seems to be much consternation about the use of captured bearings in high-end components. Something is fishy in Denmark.

DMF 

I too have rebuilt hubs and BBs with retainers, and not always in cheap bikes. I didn't see any reason to change them out (much less a need to cut them out). Just change the balls and pack. I've also seen loose balls that came out in pieces. Should I conclude then that captured bearings are better? Nope.

well biked

DMF, I've enjoyed the discussion.

DMF 

Same here.

kleng

Here are the Durace bearing retainers

Bob Dopolina 

Been away in Korea for a few days. Glad to see this discussion is still going strong.

A few points:

If you have never seen a stock hub with retainers, you haven't worked on new Campagnolo wheelsets in the last few years. All Campagnolo wheelsets and hubs use either a retainer or, in the case of lower end stuff, a cartridge.

I've opened up a hub and seen balls that have been shattered into pieces. A few of these hubs seemed to have been otherwise in good condition (prior to the failure of the bearing and resulting damage). Which brings me to my next points: Ease of assembly and what really causes damage.

First, it is clearly easier to slap in a retainer and go than it is to repack with loose balls. This doesn't mean repacking with loose balls is difficult, I mean cages are certainly faster and easier in a shop. If you want to use loose bearings during production, it is a process that can easily be automated so this argument is moot in terms of why a manufacturer would choose one over the other.

My second, and perhaps more interesting thought is about what really damages hubs. I think we can agree that a well maintained hub, regardless of loose or caged, will probably last quite a while. The real damaged is caused by contamination.

At first glance it might not seem that this would be relevant to the loose vs caged
debate but perhaps there is a difference in how contaminants are treated once they are present. So, a couple of questions; Does the increased friction between bearings, in a loose ball hub, increase the wear on the bearings and thereby increase the likelihood of material, from the bearings themselves, flaking off and becoming a contaminant? If so, will the constant mashing between bearings break down contaminants into a fine, mostly harmless grit, rendering it less damaging than if the same problem had occurred in a caged hub? Does the space between the balls, in a caged system, give contaminants a place to reside until the hub is repacked? By this I mean will they work themselves into the point of least resistance (like an eddy in a current) and ride it out?

Although I suggested that we avoid empirical evidence it might be all we are left with in this case. I think it would be interesting to open up various loose and caged hubs and see if there is a difference in the condition of the contaminants. I think that we can assume that the contaminants themselves will be a constant in that they are all randomly encountered from an outside source. What may be different, however, is if contaminants from the balls themselves are present.

I also think sample size hear really matters. To convince me either way I would need to hear from several sources who each looked at a large number of hubs from either side of the debate.

Who's game?

cny-bikeman

You don't get it - the dustcap was not a separate unit but rather one solid piece with the hub, meaning there was no way to remove the retainer from the hub (or the bearings from the retainer).

DMF 

You're right. I don't. How did they get the retainer et al. into the hub in the first place?

DMF 

Me. And I will even supply the popcorn!

cny-bikeman

From what I could see it was inserted as the hub was being manufactured. Very cheap steel hubs. Made for quite a few laughs at the shop.

Mhendricks

Here's my reasoning. On older bikes or bikes that I've sold , I've come across many a retainer that have broken off pieces from where they hold the bearing in place. In the process, the broken pieces come in contact with the cones and cut into them or caused some damage. Eliminating them eliminates the problem and I've yet to find a retainer stronger than the grade of bearings that I use.

Bob Dopolina 

So your point is that the loose vs caged debate is irrelevant in itself and that the real advantage is the removal of the cheap cage? I think we can agree that in a caged system this would be a weak point. Anyone who has wrenched has seen this failure and the resulting damage.

Can we look beyond this point and assume that the retainers we are talking about are the type being used in DA and by Campagnolo thereby eliminating this issue?

DMF 

I dunno. I'd want to see the condition of the bearings in the cases where (by implication) the retainer "failed first".