xenologer

Using a loose ball bearing square taper bottom bracket. Recently replaced the bbs (and a worn out cup).
Its possible though, since the shop I got the bb's from just keeps em all in a small drawer (you seen the kind, for rows of nuts, bolts, etc small hardware) and not in the original bottles; that perhaps I have mixed bb's from different bottles/manufacturing runs.
Heard that this might be a bad thing, since even though the bb's are the same listed size, there can be minor variations between each bottle.
So, just how bad is this situation?
Will any off-tolerance bb's simply average out over time? or will this be bad for spindle/cups?
Thanks

dabac

While mixing ball bearings from different batches isn't considered ideal, I don' think anyone can tell you exactly how bad it is, or even whether it's actually significant or not. It' vary with every instance anyhow.
The manufacturers are shooting for a specific target after all, and you may well end up with a collection that is close enough for all purposes anyhow.
When in doubt I use a decent vernier caliper, not to measure but as a kind of feeler gauge. Any two balls that are close enough in size to be pinched between the jaws of the caliper simultaneously I consider close enough for bicycle use.

sch

Vernier caliper presumed to be generic for dial caliper or digital caliper, which are much more common these days and are good enough to distinguish size differences
in the 0.001" range provided you have some experience in using them. Inexperienced userswill not do as well. Dirt cheap balls, ie ones likely sold in LBS, will be the lowest quality or next to it which is still pretty good: intra batch variance 0.001", batch to batch upto 0.005". LBS likely has mixed batch and no one will guarantee that the bottler didn't mix batches. If worried go to a bearing specialist. OTOH how good is the concentricity of the BB bearing race? No one specifies smoothness or concentricity, you just assume. Aged BB likely even more out of round or bumpy than new so likely all this theorizing is pointless especially in 60-100 rpm BB erratic cycling BB use (no one pedals absolutely smoothly).

HillRider

The "standard" for quality bearing balls for bicycle use is "Grade 25" and these are precise enough in all dimensions that mixing them from lot-to-lot should be perfectly fine. The less precise Grade 200 balls are adequate and batch mixing should be no problem either. As sch noted, bicycles aren't demanding service for any bearing.

BCRider

Bottles or bags? Different batches? PHHHHTTH! ! ! ! Old cyclist's tale obviously.

Let's look at the reality of the bearing ball maker. First off they come out of the machine in a continuous basis all done using a step in the machining and finish grinding that trips when down to size. All of these go into a big hopper for packaging. So worrying about the balls coming from a different bag or bottle is ludicrous. And if the machine is set up for the less demanding grades then any ball at any time can be anywhere within that grade's spec for size and roundness. So again thinking that you're tricking the system by ensuring balls from one package or balls from a supposedly same batch is just fairy dust.

If you want tighter ball size tolerance the way to do it is pay a little extra for a better, lower number, grade of balls. They will be both rounder and more consistent in diameter within the grade spec.

Grand Bois

You realize that you're contradicting the great Sheldon Brown, right?

dabac

I'm not saying it's common, but it CAN happen. I had some caged bearings, and doing the vernier test ones from the same cage would come out OK. But if I swapped in one from another cage it'd rattle between the jaws.

FBinNY 

Yes, it's wrong to mix ball lots.

The tolerance in bearing ball sizes is beyond what anyone can measure with a vernier or dial caliper. The dimensional standard for grade 25 balls is ±.0001", or one full order finer that the accuracy of the caliper. There is also a lot tolerance, whereby G25 balls from the same lot are kept within a finer tolerance of ±.000025". The grade number of balls is equal to the lot tolerance in millionths of an inch, so a mixed lot of grade 25 balls is no longer G25, but becomes a lot of G100 balls.

Balls are precision ground and/or lapped to size and packed by production lot for this reason. Those who handle them down stream keep them by lot, and discard the dregs of one lot before starting a new one so as to preserve the more precise lot tolerance.

In all likelihood this is more precise than needed for bicycle purposes, but if you're paying a premium for grade 25 balls you might as well get the benefit (if any). Realistically one smaller ball mixed into a bearing won't make a difference, but 1 larger one could. So I advise sticking to the best practice of not mixing lots. Not because it's absolutely necessary, but because there's no valid reason not to.

davidad

No. The tolerances of even the 200 grade balls that are often used in bike hubs are much better than the cups and cones that they ride in. I use grade 25C, but have found out that it's overkill.
https://yarchive.net/bike/ball_bearings.html

FBinNY 

It isn't a matter of absolute tolerance, and as I said grade 25 balls may be more quality than needed for a bicycle. But that doesn't justify paying for G25 and then downgrqding them to G100 or worse because of sloppy work practices.

It's also important that all the balls working within a race be closely matched, otherwise the preloaded bearing will be riding on the three largest balls, and the others won't be supporting the load equally. If only one ball is larger than the set, or if some large balls are together with smaller balls elsewhere in the set, it'll create a degree of eccentricity. This may not be a problem at the speeds bicycles operate at, but will cause vibration at higher speeds, which is why electric motors are usually fitted with G25 bearings.

No matter what grade you decide you need, there's absolutely no justification for mixing lots.

dabac

Just to be clear here: I don't use my vernier caliper for absolute measurements, I use it for comparison. Working on the assumption that the jaws are reasonably parallell, I pop two balls at at time in between them.
For what I think of as "good" balls, both will appear to be clamped "equally" hard, regardless of which goes inboard or outboard. For poor matches, one will be a lot looser, might not even make contact.
I've never bothered to bring out the micrometer to find out the size of the discrepancy I'm able to identify in this manner. I've found my comfort lever WRT accuracy, and I'm happy to leave it at that.
If you look into a bearing with dust seal removed and spot a "lazy" ball, one just being shunted around by the others, my method will pick that one out, even if the actual scale of the caliper can't.

FBinNY 

This method will be somewhat better at achieving a decently matched set, but I can't imagine why you bother. Quality balls, G25 or G100 are so cheap that not buying new (and keeping them properly sorted) just doesn't make sense.

Better matched balls lower the surface stress and wear on the races improving their service life. On most good quality cup and cone bicycle bearings that's worth far more than skimping on the balls can ever save. For want of a nail, a shoe was lost...

dabac

Well, it's in those occasions when a bike has to go back together using only the parts that are at hand at that very moment.
You know, the kind of late evening/out-of-shop kind of fiddling that sees rear cables being reused as front cables, wrenches being used as spoke nipple keys, duct tape as liner in mysteriously bulging tires ASO...
Basically doing the best of a bad situation by hardcore recycling of what's in the mixed parts bin.

FBinNY 

When working without new spares, or when shops are closed, the best batched set of balls are those that are already there. Remove them and keep the sets together, and separate from each other as you wash them, them reuse them as sets.

dabac

... assuming that they are all there....