Can't find exact headset bearing, what to do?

I need to replace the headset bearings on my bike. I have an Eddy Merckx Mourenx 69 with an FSA 1-1/8 - 1-1/2 headset.

I’ve taken the bearings out and can find a replacement for the upper headset bearing - it’s the upper bearing in this set: FSA Orbit C-40-ACB No:42 Kit

However, I cannot find a replacement for the lower bearing. The lower bearing in the set above has the correct inner and outer diameters and the height is the same, but the chamfer is different. I cannot find any kind of bearing that has the correct chamfer.

Below is a photo of the original bearing next to the one from the set. Probably hard to tell, but the original has a very small chamfer at the top and the new one’s chamfer is more pronounced.

Possibly dumb question, but what would happen if I used the wrong bearing? Specifically, what if I used the bearing from the set above?

Alternatively, how do I find the right bearing?

I can’t see how more chamfer would be a problem - I think there is just a bit of a gap then, but as long as the main surfaces are supported, it should still work fine?

Alternatively, and depending on what sort of wear is going on with that bearing, you might be able to disassembly both and re-use the outer from the old bearing with the new parts inside. However, if there is a large groove or something in the old outer, that might not work well.

  • One possible issue is that this bearing would likely sit higher up into the headtube. If so, that means the fork crown would be closer to the headtube as well. Most bikes have a fair sized gap between the fork and HT, but if it is rather close to start here, that delta could be enough to allow contact between the two which would not be good.

Is there not still a flat bit at the top of the bearing the prevents it from sitting too high?

To be honest, I don’t know. I’d probably stick it in and see if it works. :thinking:

If it’s too high, you might be able to add some sort of spacer.

Maybe for some head tubes, but many flow the chamfer vertically up the tube. And if there is a flat, what can happen in the case above is that the flat supports the load instead of the chamfer (actually meant to hold the load in most cases I have seen with a tapered connection) the headset is not properly functioning.

Note that the original on the left has no real flat top, and transitions directly to the rubber seal. This is not likely resting on any “flat” and pure angular contact from what I have seen.

Trying to find a better picture, but here is one snipped from CC:

1 Like

They are different angles, one is 36 degrees , the other 45 degrees. The bearing chamfer angle needs to match the headset race chamfer angle to seat properly. Otherwise you are asking for a whole load of trouble.

OP - On the lower bearing is it the inner or outer chamfer that is the wrong angle?

1 Like

The outer chamfer. Inner one seems fine.

Having played around with it a bit it does feel as though the new one sits higher up in the headtube so for now I’ve put the old one back in. The biggest different I can see is that it appears that the chamfer on the old one starts higher up than on the new one :person_shrugging:

That’s a shame as you’ll know that is the one that needs to match your lower headset chamfer. If it was the inner chamfer you could have just replaced the crown race with one to match the chamfer.

I ran into something similar with my Madone. It’s an older one, and apparently the bike gods at Trek decided that their standard was better than everyone else’s standard, so it’s a rather special bearing too (chamfer). I was finally able to get one on a special order from the LBS, and even they had to dig a bit for it. Trek doesn’t use that special bearing any longer, making it harder to find, and even Trek didn’t have it in stock apparently. Seems the only thing that’s standard about standards is they aren’t.

I even got one of Amazon that was ‘guaranteed to fit’ and didn’t. No surprise there, but it was a decent looking bearing.

:frowning: :cry: :sob:


I worked in IT for 40 some years. Yeah, there were often ‘standards’, and then what Microsoft got away with. They would often ‘expand’ a standard so their programs wouldn’t run on true standard compliant stuff.

And they got away with it, mostly… And HTML is a hot mess of non-standard standards. (Remember someone thought that ‘press-fit’ was better than threaded bottom brackets, and many people swap out Presta tubes for Schrader)

I bought a Lowes Kobalt brand of metric impact sockets, and the 10mm was actually more like a 9.5mm. Yeah, weird huh… So were they trying to make 9.5mm a 10mm? I don’t know how a tool OEM company could make such a boneheaded screw up! 10mm is 10mm no matter what side of an ocean you are. (One thing I didn’t check is if there is an SAE socket the same as this one. But still, an inch is an inch?

I took the bad socket back to Lowes, and had to pay for the right one! Yikes… I also had to prove the socket wasn’t a ‘true’ 10mm! WTH???

If this is the Orbit, the depth of the chamfer is specified as 3mm, which looks right for the new bearing. Maybe the old bearing was wrong (hard to tell but it does look like a 45 degree chamfer)

1 Like

I guess my point was: Keep looking, you may find the right bearing.

And curse the designer with underwear full of fire ants!!!

With physical systems, there are things you can get away with and there are things you cannot get away with. In this case, I would assume you have to get the exact match. Imagine you were searching for hub bearings, but you could only find bearings 1mm thicker than the originals. You wouldn’t be able to reassemble the hub.

1 Like

In my quest for the bearing, I found some that I was assured ‘would work’, but being a carbon frame and fork, I held out for ‘the real bearing’, and eventually found one. Yeah ‘could/would work’ if it’s not exact ‘can’ work, but in a sensitive are like the bottom bearing of a headset, maybe 'could/should/would work won’t be enough, over time. Clearances are kinda important sometimes. But if it were a beater bike or bicycle shaped object, I doubt that I’d spend a lot of time worrying about a small discrepancy in a bearing, unless…


I measured the angle on the screen and (while this is an inexact tool), it’s a lot closer to 45 degrees than 36. I say the old bearing was just manufactured wrong, not enough chamfer (height of the tool that cut it adjusted wrong maybe?). If you look in the head tube you will likely see some a ring of slight marring on the chamfered surface of the integrated cup where the bearing ended. And you will see that the chamfered surface of the cup is wide enough to match the new bearing. Thus the new bearing will work fine.

1 Like

Just to round this off, found a LBS that specialises in fixing and selling on second hand bikes. Took them all of a minute to identify the bearing as a MH P16H8.

Happily they had one in stock I could buy :smiley: