Cycling UK Forum

This is a head scratcher, rather than a problem to use, but I am wondering if others can explain ?

I've just been struggling away to get the preload ok on an old threaded headset. The headset was indexed, so I've removed the cages and installed loose balls. (now 28 balls with about half ball gap). Having tried to get the loosest setting with no movement/rock on the steering......I've noticed that whilst the forks rotate freely to 90 degree soon after they get rougher & eventually more or less bind. Tried it several times with the same result.

Obviously in use I won't get into that range of turn on the steering.............but I can't work out why it binds ? best guess is one or other of the frame fixed cup/race is not in quite straight ? ...............or would you get uneven wear front to back ?

My guess is that the headset cups aren't square to the head tube. There are some bodges to take them out and refit - but you really need the correct tools. It's a 10 minute job.

.. and are you sure the new balls are the correct size (i.e. as the caged ones)?

robgul wrote: ↑7 Sep 2021, 1:51pm My guess is that the headset cups aren't square to the head tube.

My guess too. I would surmise it has been like this from the start and the bearings have created an arc of progressively deeper wear on both cups towards the high point. As the steering is turned the wear on each cup moves towards the undamaged area, clearance is reduced and it starts to bind.

That, or a bent steerer.

If steerer is out of alignment, it'd have to be like that since I've had it.

balls are 5/32". Measured with vernier before changing, so correct I believe. ( the ones I took out may well be the original 1991 bearings...at least I can't recall changing them!! since I acquired it in 1993). Part of one cage had broken.

Certainly more wear at the front would explain why it is tighter at the back.

It's back together now, but I don't recall seeing wider tracks to the front. I'll investigate more if it comes apart again.

The binding effect is pretty symmetrical/matching on either side. With my best glasses on I can't see any mis-seating of the fixed cup/race
(although I did remove and replace them - when I did a lovely grey hammerite (!) repaint back in 2004 - using non-standard tools......carefully.....as in screwdriver as drift and blocks of wood and hammer.....tapping in various shades of ginger....checking evenness as progressed)

Whilst I don't recall any long period with a loose headset, I'm wondering whether it might be a prolonged period with too little preload with fractional movement that I didn't notice. That might fit with the "symmetrical-ness" as I presume, if loose, the wear would be at the front ?....except when breaking ?.

I'll probably be dismantling for full fettle and paint repair this winter. Sounds like best not to remove the fixed cup/race.......or if I do I'll need to make sure they go back in the same orientation.

If you are dismantling I would lash out and buy a new headset and get that fitted by an LBS with a proper press (the purists may suggest checking the headtube itself for the top and bottom to be parallel but that's seldom an issue that has any effect IME) to make sure the cups are pressed in evenly, fully and parallel.

New cups, new balls or even these new-fangled bearings should work.