This a continuance of a thread that has been bouncing around the Bicycle Mechanics Forum for a while, but I've moved to here to C&V because it affects vintage bikes more than any other. A good definition of headset indexing or fretting was posted by FBinNY I have a Nuovo Record headset that is indexing and tried a number of fixes that might be helpful to others with the same problem. First, this is on a 30 year old Italian bike that had the headset when I bought it a couple of years ago. Also, if you don't know, part of riding vintage bicycle is to keep the vintage components working as long as possible. I noticed the indexing about 6 months ago and overhauled the headset which seemed to make it better, but it returned a short time later. The indexing isn't dangerous and is unnoticeable while riding so it is an OCD matter if anything; so of course I must fix it!

The two fixes (after an overhaul) normally mentioned are to remove the ball bearing retainer and install loose bearings or to remove the crown race and turn it 90 deg (to change the position of the divots with bearings in the retainer. I first turned the crown race, but all that seemed to accomplish was to move the location where the headset indexed. This may work depending how badly the race is fretted and when I took the time to inspect better inspect the crown race I saw the fretting covered almost the entire race. I then removed the ball bearing retainer and installed what turned out to be 22 3/16" loose ball bearings (2 more than what was used with the retainer). But this really didn't work either. The fretting made it difficult to properly preset the bearings and while the bearings didn't index , they still felt rough none the same.

The last thing I tried was using the crown race off of a Tange NJS Levin, which is advertised as a Nuovo Record copy. However it turned out it wasn't a good fit; the Levin is more of a Nuovo Gran Sport copy (with the addition of the banded engraving of a Nuovo Record). The Levin uses 5/32 balls (the same as a Nuovo Gran Sport) and the Record uses 3/16". The result is the crown race is smaller in diameter, leaving a larger gap between the lip of the crown race and lower cup, which will allow more crud to get into the bearings. If I had to go this way I would probably use the lower cup of the Levin as well.

My fix finally came from ebay were Boulder Bicycle/ Rene Herse https://www.renehersestore.com/servlet/StoreFront was selling the crown race for what is most likely a Triomphe. The Triomphe and Victory headset parts are pretty much imterchangeable with Nuovo Record including 3/16' bearings and the crown race. I little history here; Triomphe and Victory were designed not only as a mid-level component but also as a legacy design to those that preferred the old Nuovo Record (Triomphe) and Super Record (Victory) over the new C-Record. In the the case of headsets Campagnolo chose not to make an alloy like the Super Record, so the Victory was made to resembled a Nuovo Record instead. To differentiate the Triomphe headset it was made without the banded engraving similar to a Nuovo Gran Sport; the Triomphe headset used the same locknut and upper cone, all other parts differed in model number. The Victory is such a close copy of the Nuovo Record that the only obvious way to tell them apart is the tool flats on the threaded top cap; Nuovo Record has 6 flats, Victory and Triomphe has 2 flats.

I have not yet received the Triomphe crown race in the mail, but I anticipate it will extend the life of my Nuovo Record headset for a while longer. I know this is incredibly long but if there is somewhere I am wrong or something I left out I'd like to hear it. Thanks

Thanks for posting, I've always just replaced the entire assembly once the "indexing" problem sets in, but have often wondered if turning the lower cup 90 or 180 out of phase might fix it. From what you reported sounds like that's not an option, so I won't waste my time trying it. Best of luck with your fix.

I would think you are going to have to replace the lower cup along with the crown race & possibly the entire headset to get rid of the indexing completely. If the headset wasn't adjusted right ( leaving it too loose is the most common thing that will cause indexing to a headset) the indexing can come from both the top & bottom cups from the consistent rocking as you ride. I don't buy the theory entirely about the grease, yes when riding the steerer isn't turning the full range but before & after riding the steerer ( at least on my bike) gets turned almost the full range a lot which should pick up fresh grease each time & if you use enough of a good quality grease it really shouldn't matter IMHO.

Glenn

Great post. And, since I recently went through this NR head set replacement thing, I find this very topical. Though I'm no engineer I might agree with Glennford about the lubrication thing. I know Jobst attributes the failure to lube failure but I would think it's mostly just caused by what you say in your thread about the balls being mostly stationary during their lives. Although in most descriptions of false brinelling I've read they do state lubrication failure as a contributing factor.
I'm also wondering if the caged Campy bearings can contribute to or exacerbate this problem. I've read theories espoused in both directions. Yes and no. In short, that even loose balls will eventually lead to this problem. Because of the design of a ball and race head set, I tend to think it'll happen no matter how many balls are used. But again, I'm just guessing here.

I removed the bearings from the cage and set them loose, rotated the cup and removed it to set them and this is how it looked. From what I saw, it looks like you can fit 3 extra bearings in there.
So, I added the three bearings and I think they fit fine. I cannot feel any indexing, no matter how tight I screw the cup down. This is without a wheel in the fork. I will retest when both wheels are installed and I can put some weight on it.
I believe this is a C-Record headset from 1987.



I've done a Photoshop mockup of this by copying 3 bearings and rotating them. they don't quite fit but there is still extra space to be gained by free space around some of the other bearings.

The fixes I stated are known fixes, but they can be mitigated by the extent of the damage done by the fretting. When I installed the loose bearings, I also did not feel any indexing, but I couldn't get the bearings to roll smoothly due to the extent of the damage to the race. FYI, the rule to go with loose bearings is all that will fit minus one; for the Record and Super Record the number should be 22.

False brinelling/ fretting is the excepted cause of indexing, not poor presetting, but you are certainly intitled your own opinion; but you should keep an open mind (here is similar definition of false brinelling https://www.engineersedge.com/lubrica...brinelling.htm). As far as the extent of the damage on the entire headset, a thorough inspection of the headset races and cups show absolutely no damage whatsoever besides the crown race, which is common with fretting. This can not happen if the headset is too loose, but is consistent with the fretting definition by FBinNY (and also from Jobst Brandt https://sheldonbrown.com/brandt/indexed-steering.html ). . It is simply a fact that the entire weight of the frame and rider rests on the crown race. Assuming false brinelling is the case, the crown race would take the brunt of the damage.

The real problem is that ball bearings are not well suited to the type of loading expected in a headset. A better solution is roller bearings, as used by Stronglight, Avocet, and others.

This is a common argument, but even roller bearings are not immune from false brinelling. Here is a photo of the roller bearings used by Stronglight and below is an A9 I used on a bike for a couple of years. As you can see there are false brinelling marks on the race; however the good thing about the A9 is the races come with the rebuild kit!

Here's mine, after 20 years with zero maintenance:



I still think roller bearings are better suited for headsets than ball bearings.

One thing you might want to be careful of, Von Carlos, is this. I've read, and heard mentioned here as well, not to fully load those loose bearings. It might look like it will take three extra balls, but they say to fully load it, then remove two. You might try to remove at least one. Three is too many I think. Too cramped. YMMV.

Still a bit of fretting damage on your lower race, JD, but I think you're right. Rollers are less susceptible to this than ball bearing headsets. Newer sealed bearings eliminate the phenomena I've heard.

True; that is because the rollers have line contact instead of point contact, so the effective unit loading is lower.

I'm skeptical; why should this be so?

Apparently it has to do with the use of Spherically Aligned Cartridge Bearings; from Jobst Brant
Huh?

I'd need to see a drawing to understand it exactly, but it sounds like they have a composite assembly based on a ball bearing race, with features to mitigate fretting: radial preload and a support bearing.
Thanks for the explanation.