Accuracy Of Bottom Brackets From Major Brands Analyzed

[waltthizzney]

Pretty Interesting Video by this guy

https://www.youtube.com/watch?v=zryhuHkbb-o

Basically what he found was Cannonade BB as expected are the worst, and Hongfu/Chinese Brands are making better units.

[bikerjulio]

I enjoyed that. Thanks.

[Calnago]

That pretty much mirrors my practical experience as well. Cannondale, despite blessing us with the wonderfully creaky BB30, only to find the shells were so bad they had to come up with plastic cups (PF30) which would "deform" a bit to match, has failed to ever get it right it seems. And now we seem stuck with it. They leave it up to the shops and individuals like me to fix what shouldn't have to be fixed. Thanks Cannondale. I have also found that their molded headset races are sometimes not very good either, leading to difficulties in proper headset bearing seatings, which results in having to apply excessive preload to the bearings which in turn results in premature wear.
The issues are not confined to Cannondale, but they sure seem to have more than their share of creaking bb's, which is ironic since they are the company that introduced the whole idea in the first place.

I was going to ask Hambini just what the sample size was for these tests, since one out of round or misaligned BB does not necessarily mean that's a consistent finding that you can extrapolate any conclusions from. But I read through some of the comments and it appears that he tested over 200 frames and had to have a sample size of at least 12 for it to be included. That's not too bad I suppose. I pretty much do the same tests that he does, but only on frames I work on, just to see what kind of tolerances I'm starting with in any new bike I build. I certainly don't have a huge sample size population of frames, but when I see a Cannondale, I generally know what I'm going to be doing to fix it in advance.

[wheelsONfire]

Incredible...
A pity there is no BB 386 (EVO) version.

[goodboyr]

That pretty much mirrors my practical experience as well. Cannondale, despite blessing us with the wonderfully creaky BB30, only to find the shells were so bad they had to come up with plastic cups (PF30) which would "deform" a bit to match, has failed to ever get it right it seems. And now we seem stuck with it. They leave it up to the shops and individuals like me to fix what shouldn't have to be fixed. Thanks Cannondale. I have also found that their molded headset races are sometimes not very good either, leading to difficulties in proper headset bearing seatings, which results in having to apply excessive preload to the bearings which in turn results in premature wear.
The issues are not confined to Cannondale, but they sure seem to have more than their share of creaking bb's, which is ironic since they are the company that introduced the whole idea in the first place.

I was going to ask Hambini just what the sample size was for these tests, since one out of round or misaligned BB does not necessarily mean that's a consistent finding that you can extrapolate any conclusions from. But I read through some of the comments and it appears that he tested over 200 frames and had to have a sample size of at least 12 for it to be included. That's not too bad I suppose. I pretty much do the same tests that he does, but only on frames I work on, just to see what kind of tolerances I'm starting with in any new bike I build. I certainly don't have a huge sample size population of frames, but when I see a Cannondale, I generally know what I'm going to be doing to fix it in advance.

+100 on this. My experience with Cannondale bb's and headsets matches yours. I would also add that trek bb90 early versions were super bad as well.

Sent from my SM-G900W8 using Tapatalk

[kkibbler]

Ironic since iirc Cannondale said they designed BB30 the way they did because they had confidence in their carbon manufacturing tolerances.

[JOrange]

Love Hambini's videos, this one's insightful, just like the one on aerodynamics

[Boshk]

I wonder if Cannondale or Specialized reps bring this up in their meetings?

[Valbrona]

The guy is measuring BB shells. Incidentally, he doesn't say how he is doing this in what seems to be his back bedroom. But measuring is not how the industry tests its BB shells because even with BB30 there is actually a tolerance range. With PF30 there is a wider tolerance range.

The industry uses 'go' and 'no go' tools. I believe FSA market these.

[darnellrm]

Go and No-go gauges are great for checking items quickly during manufacturing, but are certainly not as precise as measuring the item directly.

[pyrahna]

The guy is measuring BB shells. Incidentally, he doesn't say how he is doing this in what seems to be his back bedroom. But measuring is not how the industry tests its BB shells because even with BB30 there is actually a tolerance range. With PF30 there is a wider tolerance range.

The industry uses 'go' and 'no go' tools. I believe FSA market these.

Go/No-Go gauges are a less expensive and faster (the faster is important when you are testing 100s of frames a day) way of doing what he did. It also tells you less information. I would like to know what kind of measuring tools he used as well, but not using a go/no-go tool would be a sign of a well designed test. Any bore measuring tools or nice enough height gauges, a granite table and some good technique are plenty good for doing what he did. A 100k+ CMM would be overkill but would also do it in a more repeatable and programmable way.

I would be very very surprised if the 'Industry' doesn't use a CMM to test samples they pull out of the line for more thorough QC checks. It's pretty standard in most industries where moderately tight standards are held.

I disagree with his views on CNC machining. Good machine shops use automatic tool wear checkers that automate the process of applying a wear offset to keep parts in tolerance. I can guarantee that the races on the bearings he sells are made w/ CNC machines and then ground using an automated process. Even competent shops that don't use an automated wear offset will check the occasional part and apply an offset at the machine for a tool to keep tolerances tighter than spec'ed.

Just because he is in his back bedroom for the videos doesn't mean he doesn't have access to these types of tools. It just means he doesn't have access to a nicer video studio.

[bikerjulio]

He also commented on poor knock-off bearing tolerances, mentioning FSA also, as contributing to the overall "out of tolerance" issue.

I didn't go back and check, but the BB30 standard had published tolerances (in terms of diameter anyway), and I'd assumed this is what he was referring to in his charts.

If the "go" and "no go" used in manufacturing resulted in BB shells that were within spec, then why the results, and variation between brands?

[WinterRider]

Good machine shops .

Are 'good' when accountants do not determine the output. Isn't this why he finds all this 'crap' in the "creaking" offerings?

The checkbook buyer likes paint schemes... advertizing dung that turns them on... looking good on the road. Actual performance matters little when the bike sits mostly in a garage.......

Marketing companies are not in the business of selling quality bicycles.

[pyrahna]

Good machine shops .

Are 'good' when accountants do not determine the output. Isn't this why he finds all this 'crap' in the "creaking" offerings?

The checkbook buyer likes paint schemes... advertizing dung that turns them on... looking good on the road. Actual performance matters little when the bike sits mostly in a garage.......

Marketing companies are not in the business of selling quality bicycles.

'Good' is relative. In an earlier part of my life I worked at a company that made transmissions for race cars. The primary machine shop we relied on was what I called 'Good'...any place that held a higher tolerance I considered 'Great' and any place that couldn't hold those tolerances I considered 'Crap'. It was a pretty high standard considering that they made parts for us and parts for fighter jets. For some reason I didn't know of many shops I considered 'Great'

In all seriousness good shops give you parts that are within the specs on the blueprint, 100% of the time. They will quote you according to the specs on the blueprint and the difficulty to make the part. Usually accounting gets involved when the original quote comes back too 'high'. The accountants then ask what can we do to make this less expensive to make and the 'good' shop says move the decimal place on the tolerances....accountant says sure, no problem. Parts are then made consistently to a lower precision.

The part that I find crazy is that the cannondale and boardman whisker charts extend past the size of the bearing at the upper end. That would mean that the bearing is no longer a press fit and will just sit there. That is just insane, and would go a long way to explaining the creaking in BBs.

[bremerradkurier]

Could cryofitting be a solution to creaking pressfit BBs or are the frame tolerances too sloppy?