loose headset

[jules21]

my headset keeps loosening. it's got press-fit bearing cups and sealed cartridge bearings.

when i first installed the cups, i couldn't seem to press them in all the way, so, being a lazy sod i thought "i'll let the road do that work for me". this seems to have worked, with the cups now seated up against the head tube. so i expected to have to periodically tighten the headset as the cups pressed in, but that need hasn't abated.

it only loosens very slightly, needing maybe a 1/16th turn of the cap bolt, but it's enough to be noticeable and induce disc brake shuuder. and it happens regularly - it only takes 1-2 rides for it to loosen.

i can think of 3 possible reasons:
1. the headstem is sliding up the steerer (unlikely, as i've changed headstems, with no notable difference)
2. the star nut is being dragged up the (steel) steerer (prime suspect - cheap star nut was used)
3. the bearing cups are still being knocked in (possible, i think it's unlikely - they're pretty snug now)

any ideas? i have ordered another star nut and was thinking of pushing that in against the existing one, creating a double star nut - the bolt might be able to go through both?

[rustychisel]

are you certain the head tube was properly faced and prepared. An out of tolerance by less than a mm will create enough room for the little sucker to squirrel loose, especially if the cups can move (even imperceptibly).

A quality star nut (alloy steerer, tight?) will at least eliminate that as a possibility.

[MichaelB]

Ive had #2 happen to me on an alloy steerer. Once the top cap was popped off, it was easy to spot, and either confirm/eliminate.

[jules21]

thanks guys.

rusty - no, i have little confidence that the head tube was properly faced. this is my commuter-MTB and the frame was bought for as little as possible. the problem first presented when i fit the forks to the new frame. it's a cheap no-name job. what exactly would work loose though - the cups? they are this type:

my impression was that they're pressed on pretty tight. or is that if they are not concentric, that the bearings will have play in them?

michael - i've inspected the star nut previously and it hadn't obviously moved. i'll have another look and probably fit the replacement one i have on order (although it's a cheapie too)

[rustychisel]

cups will move around their axis to point of least resistance, ie looser, if not firm and with perpendicular faces. It can take a while, but seems more an issue for A-Head threadless headsets in my experience.

[jacks1071]

Assuming the bearings are seated and its done up properly, then the only culrpit could be the stem slipping.

Is it an alloy or a carbon steerer?

How much torque are you using on the stem bolts?

The star nut won't be the problem, in fact it should have no load on it once the stem is done up ie. you could remove the top cap and should not experience any issues. The star nut is for the initial tension only - once the steerer is done up the star nut should not be doing anything at all.

Possible solutions in the order I recommend: 1. increase stem bolt torque - An alloy steerer should be safe up to 10Nm or torque. Carbon steerers will usually take 6-8Nm but check your manufacturer's recommendation as an over-torqued carbon steerer can end in disaster, 2. use a stem with a LARGER clamping area, 3. Try some carbon grease.

[jules21]

you guys are making sense. of course the star nut doesn't affect tension *facepalm*

i will try upping the stem bolt torque. i don't have a torque wrench but i have been giving it a bit of tension - they are only dual 4mm hex bolt heads so i can't go too nuts but maybe i'll try a bit more.

it is an FSA Aheadset too (not the one in the photo). rusty, i have a feeling you could be right about the head tube facing - to eliminate the bearing slack, i do seem to need to apply a decent bit of tension to the cap bolt. if the bearing seats were machined accurately, i think they should seat without needing to be torqued down?

[drubie]

You might want to check that the strategy of riding around with non-seated cups hasn't ovalised your head tube.

I've got one frame here with an ovalised head tube and it displays the same symptoms (headset loosens over time).

Unfortunately, there aren't many fixes for an ovalised head tube - either fit an FSA "The Pig" with longer bearing cups to get past the ovalisation (or the really expensive Chris King headset with the full 1" cups) or try "staking" your existing headset cups inside the head tube by bashing the bit that doesn't quite fit with a nail punch and perhaps using a bit of loctite on it.

Just be careful with the staking procedure, because sometimes you will have fixed the loosening headset but created a binding one instead because the cups aren't quite parallel - more noticeable with cartridge bearings than ball bearings (i.e. I'd just use The Pig and be done with it).

[rustychisel]

it is an FSA Aheadset too (not the one in the photo). rusty, i have a feeling you could be right about the head tube facing - to eliminate the bearing slack, i do seem to need to apply a decent bit of tension to the cap bolt. if the bearing seats were machined accurately, i think they should seat without needing to be torqued down?

hmm, I think with angular bearings it's more likely to be an issue. Old skool ball n race bearings have a happy way of self-centreing, even though the wear rate might be accelerated. I like drubie's ideas

[Mulger bill]

Pull it down in order.
Check all parts against techdocs
Apply left palm to face when you realise you fitted the intermediate fractal thingummyring upside down.
Refit with all components correctly placed, adjust.
Ride off into the sunset...

Personal experience

[jules21]

just an update on this, if anyone's curious. thanks for all the suggestions - the winner of the meat platter is... shaun, who seems to have correctly predicted my mechanical ineptitude.

after i pulled it down and re-assembled (excluding the bearing cups, which remained), it seems OK now. it also feels different - previously i had to torque the star nut down really hard to eliminate bearing slack, but it would quickly come loose. after reassembling, i was able to adjust bearing tension more precisely and it hasn't loosened.

no idea what i did wrong with the previous assembly to cause that, but it looks like operator error

[jules21]

just an update on this but it's still not right, and i just found this:



i reckon that might be contributing to slippage

[Nobody]

Probably not. I get that too and just apply Castrol Boating grease on the outside of the steerer to prevent further rusting. If anything, rust should help with grip I would have thought.

Is the steerer just a tiny bit too small so the stem(s) can't tighten on it properly? Even with grease, mine hold well. But I do occasionally re-tighten all the stem bolts. If it's alloy stem on a steel steerer, you should be able to tighten the bolts up fairly firmly.

[il padrone]

A bit of surface corrosion will just add to the friction - should prevent any loosening of the stem.

[jules21]

i can't figure out why it's slipping

[gururug]

If you've done it a few times and it's still doing it, there is a fair chance you achieved the correct compression in one of your previous attempts.

Logically, the main cause for this would be something "uneven" in the mix. As you ride, the cup/bearing/spacer that is not square is compressing a tiny bit somewhere and driving the stem up. The stem will still be tight ( not rotatable ) but it will work its way up. Ridden like this it will wear at horizontal weak points and add to poorer tolerances if not addressed asap.

Try to find the uneven part.

[jules21]

well i've got a new frame and headset, but same problem. i reckon the disc brake shudder may be induced by wear in the suspension fork bushes, rather than slackness in the headset.

[Nobody]

Replacement fork?
http://www.aebike.com/Rigid-Mountain-Forks_c_1673.html

Please excuse me if I can't be bothered to read this whole thread again.

Is there 3 to 5mm between the top of the steerer and the top of the spacer stack or stem where the cap sits? If there is < 3mm then you might just be tightening the cap against the steerer instead of the cap sitting on the stem/spacer and pulling the steerer tight on the bearings.

[warthog1]

You were wondering if the star nut was slipping. It seems unlikely, but if so give a compession nut a go. CRC have a selection of them, you should be able to find one that fits.