Bicycle Mechanics - Bearing retainers vs Loose bearings

Hi

I was wondering which setup is more efficient and smoother running in a Dura-ace WH-7801 front hub.
The shimano spec is to use the retainers but for some strange reason my hub came with loose bearings and 2 extra bearings (to make up for the missing retainers) on each side.

14 x 5/32 bearings in a retainer or

16 x 5/32 bearings loose

Thanks

How tight are they packed in the bearing? 16 may be too many. Are the wheels new from the factory?

A bearing retainer keeps the balls separated and prevents them from rubbing against each other. It also increases the load on each ball since there are fewer of them. But it keeps them evenly spaced so the hub is more likely to run smoother, depending on the condition of the balls.

I prefer to use retainers where they're originally fitted. But I don't see a compelling reason to prefer one over the other in every situation.

How tight are they packed in the bearing? 16 may be too many. Are the wheels new from the factory?

A bearing retainer keeps the balls separated and prevents them from rubbing against each other. It also increases the load on each ball since there are fewer of them. But it keeps them evenly spaced so the hub is more likely to run smoother, depending on the condition of the balls.

I prefer to use retainers where they're originally fitted. But I don't see a compelling reason to prefer one over the other in every situation.

The wheelset was new from the factory, but I've ridden 6,500 kms on them, and this is the first time I have looked at the hubs. The 16 is n-1, there is room of another bearing in the races. Everything is in good condition except that without the retainers it isn't up the to Shimano design spec, just to be safe I'll put in some retainers which I'll need to order from Shimano. Thks

I just looked at the Shimano web site and, sure enough, they spec retainers in the WH-7800 hub. That's a first for me since I've never seen Shimano hubs of any model with retainers. If they were mine, I'd continue to forgo the retainers and just use loose balls the way your wheels were actually assembled.

Very interesting - I've only seen bearing retainers in cheap-*** hubs on cheap-*** bikes.
I'm with HillRider here - your hubs should be fine as-is with loose bearings. Nearly every other top hub (including Shimano) that doesn't use cartridge bearings has loose bearings, not in a retainer. So perhaps it doesn't matter much, as DMF says.

Very interesting - I've only seen bearing retainers in cheap-*** hubs on cheap-*** bikes...Nearly every other top hub (including Shimano) that doesn't use cartridge bearings has loose bearings, not in a retainer. So perhaps it doesn't matter much, as DMF says.

Campagnolo uses resign retainers for all hubs and wheelsets.

I just looked at the Shimano web site and, sure enough, they spec retainers in the WH-7800 hub. That's a first for me since I've never seen Shimano hubs of any model with retainers. If they were mine, I'd continue to forgo the retainers and just use loose balls the way your wheels were actually assembled.

I spoke to Shimano tech about my strange hub, and they said that there have not been any complete wheelsets in the Dura-ace range since 2005 that used or uses loose bearings. They said that factory quality control would prevent the wheelset leaving the factory without bearing retainers, so they think Giant in Australia or the LBS might have modified them.

Very interesting - I've only seen bearing retainers in cheap-*** hubs on cheap-*** bikes.
.

There's actually another level of cheapness that is lower than what you mentioned. None of the bikes I work on even have retainers in the hubs. It's all loose in there. New bikes mind you.

Campagnolo uses resign retainers for all hubs and wheelsets.
Only Campy wheelset I've worked on is 2000 Mirage that has loose balls in front hub and cartridge bearings in rear. When did Campy start doing this, and how quickly did they phase it throughout their line?

Only Campy wheelset I've worked on is 2000 Mirage that has loose balls in front hub and cartridge bearings in rear. When did Campy start doing this, and how quickly did they phase it throughout their line?

Can't remember the exact model year off the top of my head. Take a look at their parts schematics.

Interestingly the part number for all Campagnolo bearings is the same: HB-RE023 (except for the ceramic options or the cartridge bearings) which means the bearings in your Chorus hub are the same ones that are in your Record hub or your Eurus, Shamal, Hyperon, Zonda, Neutron, Proton or Bora wheelset.

Had some in my hands, today in fact.

I thought the general consensus was that loose balls was better: more weight-bearing surface, less friction. I'm in the habit of ditching any retainers I find when I overhaul.

I'd be very interested in knowing why Shimano uses a retainer only on Dura Ace hubs......or are they currently using retainers in other hubs as well?

Pulling the cages is an old practice to improve smooth running. Lets the bearing revolve without resistance.

Pulling the cages is an old practice to improve smooth running. Lets the bearing revolve without resistance.


My experience as well, and in addition I believe loose balls make bearings last longer. But again, why is Shimano using retainers in their highest end hubs, when they use loose balls in their other hubs (at least the ones I've seen)? It's also pretty well accepted that retainers make assembly easier, a major concern for manufacturers, I would think. So if Shimano is going to use retainers in their Dura Ace hubs, why not use them in ALL their hubs?

It's also pretty well accepted that retainers make assembly easier, a major concern for manufacturers, I would think.

Your answer is there, 'easability' ( is that a word:rolleyes:) plus, the amount of resistance we are talking about is and or could be, just going overboard.

Your answer is there, 'easability' ( is that a word:rolleyes:) plus, the amount of resistance we are talking about is and or could be, just going overboard.

Right, I believe the main, maybe the only, benefit of caged bearings on a bicycle is ease of assembly. But still, this doesn't answer the question: why only Dura Ace, of all groups? Or has Shimano begun using cages in all their hubs?

Only Campy wheelset I've worked on is 2000 Mirage that has loose balls in front hub and cartridge bearings in rear. When did Campy start doing this, and how quickly did they phase it throughout their line?

This came up in another thread so I looked through some of the spare parts manuals on the Campy website. It looks like they started introducing the bearing retainers around 98-99 or so with the last major redesign of the Record hubs, and eventually migrated the design to Chorus and Centaur. Veloce, Mirage, and Record Pista appear to use various other designs with either loose balls or cartridges. Now with the '07 stuff they appear to have gone to cartridge bearings on everything below Record. Record is more or less still using the classic design with the retained ball bearings and adjustable cones.

Pulling the cages is an old practice to improve smooth running. Lets the bearing revolve without resistance.

When loaded balls contact each other, there is much more friction than ball to (unloaded) carrier contact. a) the loaded balls can't easily move apart, b) the contact velocity is double that of ball to carrier.

If there is a slight size difference in the balls, they will tend to bunch behind the smaller ball and eat energy.

Right, I believe the main, maybe the only, benefit of caged bearings on a bicycle is ease of assembly. But still, this doesn't answer the question: why only Dura Ace, of all groups? Or has Shimano begun using cages in all their hubs?

They often introduce new things in DA then move them down the groups. We'll have to wait and see.

When loaded balls contact each other, there is much more friction than ball to (unloaded) carrier contact. a) the loaded balls can't easily move apart, b) the contact velocity is double that of ball to carrier.

If there is a slight size difference in the balls, they will tend to bunch behind the smaller ball and eat energy.

Which is the exact reason why you don't mix and match and use all bearings from teh same batch

This came up in another thread so I looked through some of the spare parts manuals on the Campy website. It looks like they started introducing the bearing retainers around 98-99 or so with the last major redesign of the Record hubs, and eventually migrated the design to Chorus and Centaur. Veloce, Mirage, and Record Pista appear to use various other designs with either loose balls or cartridges. Now with the '07 stuff they appear to have gone to cartridge bearings on everything below Record. Record is more or less still using the classic design with the retained ball bearings and adjustable cones.

Actually Chorus is still a bearing with retainer as well. Plus, if you look at their wheelsets, as I mentioned, most of them use the same bearing with resign retainer. The lower end wheelsets (Khamsin, Vento and Scirocco) use cartridge bearings.

This point of loose vs caged came up in anther thread recently but no-one ran with it. I mentioned there that I used to use loose bearings exclusively but now am not so sure. My reasons for this are twofold. First, I find when I am adjusting a loose bearing hub, I can get it to run very well but I can also find a spot (briefly) where there is play. This is pretty conclusive evidence (of the obvious assumption) that they are moving around in there.

This brings me to the second consideration. If they are moving around so much, they must be rubbing against one and other. This would increase friction in the system. It is much like rubbing tires when you are riding with someone; Even though you are traveling in the same direction the back of the rear wheel is going up and the front of the front wheel is going down. Bring these two together and you have friction. So if you reduce friction by adding more bearings (as loose bearings) are the gains made here offset by the increased friction of the balls rubbing against each other?

Throw in a retainer. How much friction is there between the retainer and the bearing? If you use a resign one, like Campagnolo does, I'm sure this is reduced (as well as the potential wear or damage caused by the retainer itself) but again, by how much?

Considering that both Shimano and Campagnolo use retainers in their top o' the line products may be a clue. This could also relate to production where a retainer is much faster to use during assembly (but if you've every seem women on an assembly line lace a wheel you could certainly overcome this if it was worth it).

As both companies spend considerable time and money on R&D, it is unlikely they will be forthcoming with any hard numbers. Has anyone else out there got any real data on this issue. Anecdotal evidence like' "I been running such and such in my such and such for years and it works fine." doesn't really cut it.

Any rocket surgeons out there who can help?

Only Campy wheelset I've worked on is 2000 Mirage that has loose balls in front hub and cartridge bearings in rear. When did Campy start doing this, and how quickly did they phase it throughout their line?
To my surprise, I checked the owner's manuals for my '06 Chorus hubs and both the front and rear hub illustrations show the bearings in retainers. I haven't had these hubs apart yet but that's the way the book shows them. The books cover Record, Chorus and "Daytona" and were copywrited in 1999 so the name change to Centaur apparently hadn't occured yet.

So, it seems the change occured by the end of the '90's, at least for the upper line hubs.

I doubt if any significant data exits that shows a real world difference in performance or durability between caged and loose balls. Follow your heart.

I doubt if any significant data exits that shows a real world difference in performance or durability between caged and loose balls. Follow your heart.

That's my suspicion,too but you never know. There are a lot of guys out there with time on their hands...

I'm wondering if it doesn't have a lot to do with the design of the specific retainer. Scientific or not, there's an awful lot of anecdotal evidence that tossing the retainer on the average old headset or cup and cone bottom bracket and repacking with loose balls is an improvement in most every way (except ease of assembly). Heck, if I can see, feel and even hear the difference just by spinning the part, that's pretty strong evidence as far as I'm concerned.

But, for instance, Bob, what the heck is a "resign" retainer, like you mention with the Campy hubs? See, I think when we're talking about these Dura Ace and Campy hubs with retainers, we may be talking about a whole 'nuther animal than the average old headset or bottom bracket as far as the actual retainers. I know that in industrial applications, too, there are specially engineered retainers that are there to increase performance and service life of the bearing assembly, and I'm wondering if the retainers on these quality hubs aren't more like that than the old retainers most of us think of when we're talking about retainers on bicycles.

This is just a thought, though, because I've never even knowingly seen a bicycle hub with bearing retainers in it, so maybe the retainers in these hubs aren't all that different than the retainers in a $15 headset. So what are the retainers like in these quality hubs?

I'm wondering if it doesn't have a lot to do with the design of the specific retainer. Scientific or not, there's an awful lot of anecdotal evidence that tossing the retainer on the average old headset or cup and cone bottom bracket and repacking with loose balls is an improvement in most every way (except ease of assembly). Heck, if I can see, feel and even hear the difference just by spinning the part, that's pretty strong evidence as far as I'm concerned.

But, for instance, Bob, what the heck is a "resign" retainer? See, I think when we're talking about these Dura Ace and Campy hubs with retainers, we may be talking about a whole 'nuther animal than the average old headset or bottom bracket as far as the actual retainers. I know that in industrial applications, too, there are specially engineered retainers that are there to increase performance and service life of the bearing assembly, and I'm wondering if the retainers on these quality hubs aren't more like that than the old retainers most of us think of when we're talking about retainers on bicycles.

This is just a thought, though, because I've never even knowingly seen a bicycle hub with bearing retainers in it, so maybe the retainers in these hubs aren't all that different than the retainers in a $15 headset. So what are the retainers like in these quality hubs?

Sorry, Resin.:D

I think, in a headset, loose bearings make sense. The headset isn't spinning around like a hub or BB. A different application so I think a different set of considerations apply.

The Campagnolo retainers or a very soft material that is flexible. I'm certain that it couldn't possible add wear to a bearing. Haven't seen a DA one in a while so I can't comment here.

The Campagnolo retainers or a very soft material that is flexible. I'm certain that it couldn't possible add wear to a bearing.

That's funny about the "resign" retainers, you had me really wondering.:D But seriously, since you say the retainers in the Campy hubs are flexible, that right there is a significant difference from the typical retainers most of us think of regarding bikes, the ones that seem to be more or less made of tin-

That's funny about the "resign" retainers, you had me really wondering.:D But seriously, since you say the retainers in the Campy hubs are flexible, that right there is a significant difference from the typical retainers most of us think of regarding bikes, the ones that seem to be more or less made of tin-

The Shimano Dura-ace ones are plastic and flexible as well, there is not much to them, its more like a netting to hold the bearings, nothing like the tin version

I think, in a headset, loose bearings make sense. The headset isn't spinning around like a hub or BB. A different application so I think a different set of considerations apply.
That's right, headsets are a very different mechanical situation and much more static so the larger the number of contact points to distribute the load, the better.


The Campagnolo retainers or a very soft material that is flexible. I'm certain that it couldn't possible add wear to a bearing. Haven't seen a DA one in a while so I can't comment here.
I'm glad this came up since I haven't had my new Campy hubs apart yet and I would have expected the old-type metal retainers. It's nice to know what to look for.

I think, in a headset, loose bearings make sense. The headset isn't spinning around like a hub or BB. A different application so I think a different set of considerations apply.


Certainly the most noticeable place you can improve your situtation by tossing a cheap retainer is a headset, but I mentioned old cup and cone bottom brackets as well. Granted, unless you're doing a restoration, it makes more sense these days to simply replace an old bb like that with a cartridge unit. But since we're talking "theory" here, I do wonder what the real differences are in terms of performance and durability, with caged vs. loose balls, in something like an old cup and cone bottom bracket with one of the old "tin" cages. Or for that matter, what the differences would be if you used "tin" cages in a hub, vs. loose balls.

My experience has been that an old bottom bracket at least seems smoother with loose balls, and it also makes sense that the most important part of the unit, the cups, would last longer with the load spread out over more area as with loose balls. This is definitely the case with headsets, but again, I'm thinking it probably is still best to use loose balls vs. cages no matter where on the bike you're talking about, if the cages are the cheap tin type. Otherwise, why would any manufacturer have EVER used loose balls in hubs? Seems to me it would be cheaper to manufacture the hubs with caged balls, and if the performance is actually better with caged balls, there would be no reason whatsoever to EVER use loose balls. Again, I think it's likely that the key factor with these quality hubs (Campy, Dura-Ace) is the type of cages they're using-

Which is the exact reason why you don't mix and match and use all bearings from teh same batch

True. Yet even within the same batch there are variations. All you can do it keep them to a minimum.

the most important part of the unit, the cups, would last longer with the load spread out over more area as with loose balls.

If ball-to-ball friction produces residue, that would effect the cups too.

Seems to me that in a radial bearing peak load occurs on one ball at a time, and that having the balls closer together does not reduce peak load, but reduces the time that load is applied. If degradation phenomena such as micro-welding are primarily a function of load rather than time, cup degradation should be unaffected by relatively small differences in ball count.

If ball-to-ball friction produces residue, that would effect the cups too.

Seems to me that peak load occurs on one ball at a time, and that having the balls closer together does not reduce peak load, but reduces the time that load is applied. If degradation phenomena such as micro-welding are primarily a function of load rather than time, cup degradation should be unaffected by relatively small differences in ball count.


I can tell you that a headset with loose balls will most definitely last longer than a headset with caged balls, all other things being equal. Headsets are different than hubs/bottom brackets, yes. But it's interesting that on older, quality bikes, before cartridge bearing bb's and the like, the places that caged balls were commonly used are also the most difficult and time consuming places to pack loose balls: the headset and bottom bracket. I think it was to reduce assembly time, nothing more. Quality hubs always used loose balls.

Now with these "higher tech" cages that are being used with Dura Ace hubs, etc., it seems likely that the benefits you mention can be put to use without the drawbacks of cheap retainers. To be clear, I have no scientific evidence of this of course, it's all speculation on my part, but after packing a lot of ball bearing assemblies over the years on bicycles, there's no way I'll be convinced that an assembly with a cheap "tin" cage as a retainer will spin with less friction than an assembly with loose balls. I'll concede that the durability issue may be a wash with something like a cup and cone bottom bracket, because a quality bottom bracket has very durable cups, likely to last a very long time regardless. But you can see, feel, and even hear the difference in terms of smooth operation-

Agree on headsets. I was talking about a radial bearing (where the load is perpendicular to the axis of rotation) and amended my post to be clearer.

To further add to the confusion, I just checked Shimano's website and looked at the exploded view of the Dura Ace hubs, front and rear. The front is apparently available with two types of bearing assemblies, an A type or a B type. The A type uses loose balls, the B type uses the retainers. On the rear hubs, it's loose balls only, no "B type" available. It sounds like kleng's Dura Ace front hub is an A type (w/loose balls from the factory), and that's why there were loose balls in it-

To further add to the confusion, I just checked Shimano's website and looked at the exploded view of the Dura Ace hubs, front and rear. The front is apparently available with two types of bearing assemblies, an A type or a B type. The A type uses loose balls, the B type uses the retainers. On the rear hubs, it's loose balls only, no "B type" available. It sounds like maybe kleng's Dura Ace front hub is an A type (w/loose balls from the factory), and that's why there were loose balls in it-


I spoke to Shimano tech about my strange hub, and they said that there have not been any complete wheelsets in the Dura-ace range since 2005 that used or uses loose bearings. They said that factory quality control would prevent the wheelset leaving the factory without bearing retainers, so they think Giant in Australia or the LBS might have modified them.

His hub is on a 7801 wheelset.

His hub is on a 7801 wheelset.

Ah, I didn't catch that, thanks-

Check out this web page.

http://www.bardenbearings.co.uk/index.cfm?articleid=635

Basically it says the load carrying abililty is greater without the retainer but "With the retainer removed from the bearing the balls are no longer spaced apart and can make contact with each other. With the inner ring acting as a sun gear each ball is driven in the same clockwise direction. At the points of contact the balls are travelling in opposing directions leading to friction and wear. Due to the introduction of ball to ball friction the bearing torque is also increased. The use of ceramic balls can provide lower wear levels and reduced bearing friction/torque."

When loaded balls contact each other, there is much more friction than ball to (unloaded) carrier contact. a) the loaded balls can't easily move apart, b) the contact velocity is double that of ball to carrier.

If there is a slight size difference in the balls, they will tend to bunch behind the smaller ball and eat energy.

I guess nobody read this. That would be a reason to only use carriers in dura-ace. It's less friction

Check out this web page.

http://www.bardenbearings.co.uk/index.cfm?articleid=635

Basically it says the load carrying abililty is greater without the retainer but "With the retainer removed from the bearing the balls are no longer spaced apart and can make contact with each other. With the inner ring acting as a sun gear each ball is driven in the same clockwise direction. At the points of contact the balls are travelling in opposing directions leading to friction and wear. Due to the introduction of ball to ball friction the bearing torque is also increased. The use of ceramic balls can provide lower wear levels and reduced bearing friction/torque."

That's interesting. It seems the primary purpose of their design, which is without a retainer, is to increase the carrying capacity of the bearing, whether axial or radial, and they acknowledge there is increased friction and torque because of the removal of the retainer. But I still say, in regard to reducing friction/torque, that it would greatly depend on the design of the retainer, too, because the typical "tin" retainers found in bicycle bearings don't strike me as being very well engineered for reducing friction. I'd bet the retainers in the Dura Ace and Campy hubs are actually designed to do some good regarding performance, and aren't just there to speed up assembly-

Check out this web page.

http://www.bardenbearings.co.uk/index.cfm?articleid=635

Basically it says the load carrying abililty is greater without the retainer but "With the retainer removed from the bearing the balls are no longer spaced apart and can make contact with each other. With the inner ring acting as a sun gear each ball is driven in the same clockwise direction. At the points of contact the balls are travelling in opposing directions leading to friction and wear. Due to the introduction of ball to ball friction the bearing torque is also increased. The use of ceramic balls can provide lower wear levels and reduced bearing friction/torque."
Even if this is theoretically correct, it's still an "angels dancing on the head of a pin" type of argument.

Loose ball hubs, properly lubed and adjusted, have such a small amount of rolling resistance that any "improvement" is almost unmeasurable. Also, properly adjusted hubs show almost no wear of either the cones, races or balls for thousand and thousands of miles. I have a set of older 9-speed Dura Ace loose ball hubs with 45,000 miles on the original cones and races and they are overhauled only every 5000 to 6000 miles. Obviously the "wear and friction" have to be very low.

These are the points that make ceramic bearings a costly extravagance with no demonstrable benefit except to the seller.

Still, it knocks the argument about loose balls being better into a cocked hat.


And it's "angels" that have rhythm, not "angles".

Still, it knocks the argument about loose balls being better into a cocked hat.

I don't think it does at all. Keep in mind the link is to a company's website that makes bearings with loose balls, and touts them as being able to carry a greater load because of it, axial or radial. And again, why would Shimano make hubs with loose balls if they didn't feel like there are performance advantages? Are the retainers in the Dura Ace wheelsets, the ones kleng describes as being kind of like a netting and nothing like the old tin retainers that are most commonly seen on bikes, designed in such a way that they do create some advantages? I think they probably are, but I seriously doubt that all retainers are created equally in terms of enhancing performance, and that these netting-like retainers on the Dura Ace wheelsets are worlds better than a retainer you'd typically see in a bicycle bearing.

So if you're saying that these netting-like retainers (and probably the ones on the Campy hubs, too) on these high quality hubs create some performance advantage, I agree that the answer is probably yes. But if you're saying that a bearing on a bicycle with the balls retained in a typical tin cage is better than loose balls, I disagree.

Btw, apparently the only hubs Shimano makes with the retainers are on the Dura Ace wheelsets and the Dura Ace front hubs, model B. All of their other hubs use loose balls, including the Dura Ace "model A" front hubs, and the rear Dura Ace hubs-

I thought the general consensus was that loose balls was better: more weight-bearing surface, less friction. I'm in the habit of ditching any retainers I find when I overhaul.

+1 Yes, loose balls are better and when replacing it's been my practice to fill all hubs and headsets with balls then minus 1. Also use only grade 25.

To further add to the confusion, I just checked Shimano's website and looked at the exploded view of the Dura Ace hubs, front and rear. The front is apparently available with two types of bearing assemblies, an A type or a B type. The A type uses loose balls, the B type uses the retainers. On the rear hubs, it's loose balls only, no "B type" available. It sounds like kleng's Dura Ace front hub is an A type (w/loose balls from the factory), and that's why there were loose balls in it-

Thanks for that, I was able to confirm Shimano directly that my Shimano WH-7801 wheelset front hubs should have retainers in them.

When I saw the loose bearings in the hub I thought I had a WH-7800 hub which had 14 x 3/16 loose bearings, but the funny thing was that the loose bearings were 5/32, so when I put in 14 3/16 bearings they would not fit and allow the cone to screw onto the axle. I was able to deduce that I indeed have a WH 7801 which can only handle the 5/32 bearing size and that the missing retainers were compensated by having an extra 2 bearing in each side.

And it's "angels" that have rhythm, not "angles".
Right. I noticed the error when I reviewed my posting but spelling was never my strength. I'm glad the meaning was clear even if the spelling wasn't. I just edited the post to correct it.

+1 Yes, loose balls are better

You people keep repeating that. I have yet to hear any reasoning supporting it. Is it because your grandfather told you so?

You people keep repeating that. I have yet to hear any reasoning supporting it. Is it because your grandfather told you so?


I gave you some reasoning, the best that I could do, which was based on experience.

Take an old cup and cone bottom bracket with caged bearings, pack it with fresh grease, and spin it. Listen to it as you spin the cranks. Feel it as you spin the cranks. Spin the cranks and see how long the cranks spin after you let go. Then take it apart and repack it with loose balls and fresh grease, and repeat the test. Or just do it on another bike, with a similar bottom bracket, if you've got one. Then get back to us and see what conclusion you reach. The difference with headsets, as already agreed upon, is even more noticeable.

I've never heard anyone say that they repacked a (fill in whatever part here) with retainers on a bicycle, and then say it works smoother and quieter than it did with loose balls. But I've heard the opposite quite a lot. Why do you think that's so? And by the way, my grandfathers were great guys, but neither one of them worked on bikes as far as I know, and they never advised me on them. Nope, I just used my eyes, ears, and hands to reach the conclusion I have regarding ball bearing assemblies on bicycles-

(Not talking about headsets. I grant you that one.)

With all due respect, you're not reasoning. You're extrapolating from subjective experience. (Nor is your experience universal. Mr. Dopolina has related that his experience with hubs is different.)

Further, your experience is of an extremely limited subset of the total functional life of the bearing — typically only the first few rotations.

However, the scientific method is partly based on experimentation, so experience has a place. We just have to be careful how we use it. Try this.

I challenge anyone to feel or listen to a functional hub or BB, loaded or unloaded, after 1000 mi. of use and tell whether it has loose balls or a retainer. You should not know what type of bike or hub/BB it is. For statistical significance*, you have to be right (or wrong!) about 9 times out of 10.


*And technically, the result must be adjusted by the actual occurrence of each type in the tested population. E.g. if 90% of the samples are of one type, then a random response yields 90% accuracy. So 99% accuracy would be required.

I would find it hard to believe that the difference in efficiency/drag would be significant enough to choose one over the other. If the cage is the type that has a full complement of ball bearings I would go with that for the ease of maintenance mainly and the theoretical reduced friction as a side benefit. Otherwise I would choose loose balls because the added contact surface seems to promote smoother bearing wear. Rough bearings would certainly pose a greater friction cost than loose vs. caged, and most people do not check their bearings very often. Certainly in a shop setting I would choose loose balls (or a full complement cage) exclusively so as to give the customer the longest lasting overhaul.

p.s. My assumption of a small efficiency difference may be wrong, but I have not yet seen a reference to an actual measurement of a difference.