jonwvara 

I've read that ball bearings of a given size and type shouldn't be mixed from one batch to another--the theory being that sub-microscopic differences in manufacturing tolerances potentially mean that the ball bearings in one batch could differ a tiny bit in size from an earlier or later manufacturing run. This presupposes, I guess, that Wheels Manufacturing or--whoever it was that poured the ball bearings into that handly plastic bottle and sold it to you--was conscientious about not mixing one manufacturing lot with another.

I'm not an engineer or a materials expert, but this has always seemed suspect to me. In the first place, I doubt that the size variation from one lot of Type 25 3/16" ball bearings to the next is large enough to matter--if indeed it's even large enough to measure. And in the second place, I doubt that manufacturers faithfully keep individual manufacturing lots separate all the way down through the supply chain.

And yet: I still have a plastic bag with four unused 1/4" balls that I've been sitting on for years because I'm apparently not sure enough of myself to go ahead and mix them with other bearings. Will someone who knows please tell me if it's okay to go ahead and mix them? Will doing so cause my bottom bracket to overheat and seize up? Should I stop hoarding them and use them for slingshot ammunition instead?

zukahn1 

High quality ball bearings are close enough in tolerence that you should fine using them. Also Campy is the only manufacturer to definetively sell bearings in matched sets.

jwarner

I don't think it would matter in all but very few very demanding applications, and I can't think of an example of one. I do know anything on a bicycle isn't one of those demanding applications, so I would go ahead and used them. My $0.02.

dddd

When you look at the actual specifications of the very good (and common and pretty inexpensive) Gr25 ball bearings, the tolerance for diameter is quite rigid, such that the difference from batch to batch is but a small fraction of the actual flex of the bearing components as installed.

So worry not about different batches, as long as the bearings supplied are actually Gr25.

The Gr25 bearings can actually make a very noticeable difference in bearing smoothness as compared to vintage bearings from the 1970's and earlier.
Today's common, low-cost Joytech hubs for example are of a much higher grade from those of the distant past used in inexpensive components.

I rebuilt a steel-bodied rear hub today from a department-store dual-suspension bike, and aside from literally having to painstakingly shim the fit of the driveside cup in the steel hubshell (.015", whew), the bearing smoothness was impressive despite the usual slightly-bent axle!

zukahn1 

Also if it helps put you mind at rest Grade 25 bearings are within 0.000025mm of spec.
https://americandad.biz/gradechart.htm

turky lurkey

I have read the same thing, and have also been suspect about mixing bearings being a big problem. I really don't think it matters. I mean how precise are we getting the preload on the bearings when we install them? I pretty much just tighten the cones until it feels about right and going by feel there is no way I am getting even close to being within a thousandth of a degree of some perfect torque, so why would I worry about bearings that are 0.000025mm or whatever difference. Even if the difference is significantly greater than that, oh well. How thick is the layer of grease between the bearing and the cup (I would think it varies from one bearing to the next at any given time)? I install the bearings with my dirty hands, I am sure there are minerals or contaminants on my hand getting in the grease. I'm no expert though, just speculating but I don't worry about it.

dddd

Worry less and live longer.

Buy Gr25 bearings in the 500 count bag and suddenly they are cheap!

I mentioned 70's-era bearings in low-cost bikes being relatively rough; ...the exception being Schwinn, who used high-grade chromium bearing balls in even their lowest-price bikes.

Twist open the ball bearing pulleys on their Allvit-derived rear derailers though, and you'll find room for major improvement upon the French 1/8" bearing balls.

old's'cool

A bigger concern with used ball bearings, especially if their provenance is not known, is not knowing how close they are to end of life. When servicing a bearing set, if there is cause to reject any of the balls, I would replace the entire set with new, following the good advice given above about choosing Gr25.

rootboy 

"Should I stop hoarding them and use them for slingshot ammunition instead?"

Yes. Unless you're aiming at birds.

Velognome 

When I buy balls from my LBS, he pours them out of a bag of thousands. Somehow the the left over don't seem to be any different then the ones I'll get the next time....that bag looks years old.

In the grand picture....how big is a lot or run? To be safe, if you could measure the difference, I'd discard; if not, I'd use them.

qcpmsame 

As my late Father-in-law used to tell us on the construction job sites, "you ain't building no Swiss watch, boy". With the production tolerances for ball bearings, unless they are at the end of their service life, as said above by old's cool, you won't be able to notice the differences, even measuring them for the differences would be difficult for the average Joe/Josephine. A machinist could of course measure them with a micrometer, easily, but most don't own one (yes, I have one, I'm an engineer/severe gear-head.) so, I'm in agreement with everyone else, use them and don't sweat it.

Bill

BTW, I had this mental image when I read the title line, someone that had rebuilt their bicycle and wanted to know how come so many ball bearings were left over, and what to do with them, now.

rootboy 

I have one too, Bill. A couple in fact. But I daresay I couldn't measure down to .000025 with it.
Can you? If so, you're a better man than I am Gunga Din. Or, have much better eyes.

qcpmsame 

It was tongue in cheek, Scott, sorry for not using the smilies. With my poor eye sight I couldn't measure a foot with a 12" ruler, mem sahib, even digital mics would give me a hard time right now. Its beyond time for new eye glasses Rx.

My mic was a gift from a friend as a joke about motocrossers and their lack of maintenance skills, it is an old screw type that has enough dust on it to make any measurements worthless. Is this enough of the smilies yet?

And, please don't tell me you think your are normal, or one of the regular types....

Bill

SJX426 

I think most screw guage mics are only good for 4 places. We are talking 6 here. What is the cost of a mike that will measure to 6 places? $1500?

auchencrow

I read on Wikipedia that the final finishing process that puts the shine on ball bearings brings them in to "between gr 10 and 48".

Seeing as how most of our micrometers can't measure that finely, does anyone really know if there is an actual difference between the premium gr 25 balls and the run of the mill shiny balls?

rootboy 

Ha. I must be sub-normal. I had to actually sit down and think, just how many zeros is that?
Let's see, .000025 is equal to 0.64 Micrometers. Or 635 Nanometers.

heck…I don't even know how much a whole lotta 9s is.

I think Auchencrow is on to something. Who's to say we're actually getting Grade 25 balls!?
I smell a conspiracy here.

Amesja

But The Sheldon said...


Just use the balls and mix leftover new ones with the new ones as they get used. I don't mix and match used with new ever. I rarely will re-use the balls as they usually aren't worth the trouble cleaning and re-using even if they are are good. They are so cheap -even the good grade-25 ones from Wheels Mfg. I use on decent bikes. I tend to use cheap Pyramid grade-200 ones on cheap vintage bikes, like Hillary Clinton said, "What DIFFERENCE does it make???"

As for what to do with the old balls. I like to keep the 1/4" ones at least, as they make nice bouncers to stick to the magnet on the handlebars of my motorcycle for tailgaters. Just reach over and grab a handful off the magnet and drop straight down. They make an awful racket when they bounce back up and hit the underbody of a car if it is following too close. I don't do this on the bicycle as I don't have a 100hp engine on tap to lose any aggro motorists.

himespau 

What I'd always heard and followed was not mixing used and new not not mixing between batches.

Amesja

The Sheldon says:

I agree, if we were building jet-turbine engines that were spinning at 15,000RPM then we would really need to watch bearing tolerances.

But the simple fact is that on bicycles it doesn't really matter that much. Most people don't even properly face BB and Head tubes so that everything lines up straight to begin with. If the races aren't perfectly straight then it doesn't matter if the bearings are perfect. Lots of folks jam wheels with the wrong OLD into dropouts that are not spaced and/or aligned properly which causes the axle to bend very slightly, also causing the races to not align in the same plane. These errors are much larger than what the difference is between production runs of grade-25 bearings...

Worry about the BIG things first before stressing about the tiny things.

SJX426 

@Amesja - nice reference. I now have thought about using the balls for red light runners. Just toss them when the light turns red!

There is a cross walk I have to use commuting home that crosses the George Washingting Parkway by the Wilson bridge in Alexandria. Last week there were two of us bikers waiting to for the walk signal (leagal to use the walk for bikes - part of the MUP). When it turned and the last red light runner passed, I went across. The other guy waited. While I was 3/4 across the street, a car northbound ran right through the red light. If the other guy had followed me, he would have been hit. The driver didn't even touch his brakes! Close one.

BTW: Thinking is different than doing.

rootboy 

I agree, Amesja. And Sheldon never mentions how on earth we are to determine which production batch they came from, anyway.

How much does ultra precision matter in a bicycle? Not all that much, I suspect. But high quality parts do seem to run better, to me. So maybe precision to a certain level matters. But, as has been said, we're not talking Swiss watches here. Or turbine engines. Or F1 engines. Or any engines. Or wheels on a Fiat, for that matter.

SJX426 

@rootboy - Wheels on a Fiat? I smell a story!

Amesja

Balls always wear out faster than the races. It's been a long time since I looked into why but it has something to do with the fact their surfaces are more tightly curved since they are smaller as opposed to the more flatter curve of the races since they are larger over-all. They are smaller and spinning faster and have less surface area to contact. But I'm not an engineer, so don't quote me on that one.

fietsbob

bag of 1000 bearing balls is cheap enough to use new ones every time you clean, inspect the races, and re grease the part.

rootboy 

Nah. Just picked one out of the air. Could've been a Volkswagen. Or a Yugo. Did Yugo use bearings in their wheels?