Headset loose straight but tight when steering turned 90*

[BALBAL]

I have a new steel 1" threaded headset between a Fort alu frame and some 753 steel forks (1" ISO with 24 threads per inch). The local bike shop fitted the headset and they faced the frame. However when the steering is set straight the headset is loose and when the steering is turned 90 degrees in either direction it tightens up to the right loading. The headset is new and a mid range Tange with caged balls (not cartridge). I have used them before and they are generally excellent. The headset is the right sizing/threading for frame and forks.

The local bike shop did not do any work on the forks. I went in afterwards and they seemed to not have the necessary kit to the face the crown race facing on the fork.

Any advice?

Different bike shop?
Go to a proper frame builder?
Fit a cartridge headset instead?

Any ideas to the cause?

[hamster]

Either the cup or crown race is not in straight or it's ovalised. Whatever, it goes back to the workshop that fitted it for re-fitting...but as a system. It's impossible to guarantee it's right without assembly.

[meic]

Have you looked around where the headset and headtube meet? Is there any gap no matter how tiny?
You may be lucky and all that is needed is a tap or two with a hammer through a block of wood on the headset to protect it.
I received one like that from Fatbirds, with exactly the same symptoms.

[Valbrona]

Sounds like they did not even have a proper crown race setter to fit the crown race properly.

[Brucey]

something isn't straight, I'd have said. Probably the crown race, although it could be the adjusting race too. Similarly they may have faced the frame but that is not at all the same thing as fitting the set races properly.

If you have spent years messing about with this stuff the MK1 eyeball can tell you plenty. I can tell you what to do here, but if you don't have an eye for it you won't see it.

Test 1; turn the fork whilst eyeing a fixed point on the edge of the lower fixed cup. You should see if the crown race is swashing at all. If it is, the crown race needs resetting, probably the crown needs a tickle with the facing tool too.

Test 2. Repeat test1 for the upper races. You should see if the adjusting race is running true or not. If it isn't, most likely the threads are badly cut on the steerer. Sometimes they will come good if a better chaser (i.e. one that locates well down on the steerer so that it can't cut cocked) is used, but if the thread is bad enough nothing will fix it. This would be a bad outcome for a 753 fork.

Test 3. Eye a fixed point on the crown race and turn the fork. If the axial clearance from your chosen point to the lower fixed cup varies, the fixed cup is cocked. More frame prep required.

Test 4. Repeat of test 3 for the upper races. If the upper set race is cocked, more frame prep required.

Tests 1 and 2 are easy; you can spin the fork 360 degrees if there is no brake, wheel and handlebars attached and swash should be obvious. Tests 3 and 4 are less easy to do. However if you can lay your hands on a set of Vernier calipers that will open up wide enough, you can measure the 'length of the head tube' (set races and all) in various positions and thus determine if the set races are cocked or not. [In theory if both races are cocked identically, this test will fail; however the chances of this are slight.]

If (say) the crown race is a little cocked (which BTW could be because the steerer is bent so that the crown seating doesn't appear to be angled to the crown, and appears to be faced correctly, but the whole crown is angled WRT the upper part of the steerer) the balls won't sit between the races in the right way. Typically this makes the headset loose one way and snug the other. However once the steering is turned, the balls roll in a new path that doesn't coincide with their previous location on either race, and the result is almost invariably binding.

Quite often if you (say) turn the handlebars 90 degrees in a single movement, the steering will bind. If you then work the steering back and forth (by about ten degrees each way) a few times, the binding will often mysteriously ease itself somewhat. This is because the balls can settle in the new position. If this doesn't happen, it can mean that the ballraces are already slightly marked up. It could instead mean that both the moving race and its fixed partner (at one end of the head tube) are cocked in the same way, so that when the steering is turned, there is a powerful wedging action.

FWIW if a headset like a Tange Levin is fitted correctly, (ideally using loose balls) it ought to comfortably outlast the rest of the bike. I have yet to be able to say that about a similarly priced cartridge bearing headset, or one costing several times as much, come to that.

cheers

[NATURAL ANKLING]

Hi,
In theory if one of the four bearing parts is, to rob a word cocked, then it wont feel good but it should not wedge and go tight.
So there must be two parts that are not correctly square and or not concentric ' worn / damaged, which might be and I have seen one of these, poorly made bearing part where the mounting hole is offset, and yes it was a cheap part.
And one or more of the parts could be worn already (second hand).
Another possibility is that the cup has had a ding and is binding on a moving part of the steerer some how.

If there was a weight on the wheel then because the load on the bearings is not pure axial meaning its at an angle to the steerer tube, then one badly fitted / defective part would mean the high spot on race would resist turning.
But the load would need to be high and it would not probably be noticed in riding because of small angle of turn in normal riding.

Even cheap and nasty stuff can give good service.

So I would take it back to shop and say its not right, they would of known this if it was inspected and tested it by spinning the fork.
Don't be fobbed of by something like it will bed in by the shop, its true that as said it could well free up, but you weren't given that instruction by the shop were you, always sounds bad.

Unless you fancy stripping it yourself you will have to get the shop to fix, always a dilemma I know to complain.

When its apart it might well be horrible, bad when you do it yourself but fuming when you pay for it.

[fausto99]

....a fixed point on the edge of the lower fixed cup. You should see if the crown race is swashing at all.....cheers

"swashing"? Was this a spellchecker error or is it a technical term I'm not familiar with?

[Brucey]

as in 'swash-plate' for example.

cheers

[Gattonero]

I have a new steel 1" threaded headset between a Fort alu frame and some 753 steel forks (1" ISO with 24 threads per inch). The local bike shop fitted the headset and they faced the frame. However when the steering is set straight the headset is loose and when the steering is turned 90 degrees in either direction it tightens up to the right loading. The headset is new and a mid range Tange with caged balls (not cartridge). I have used them before and they are generally excellent. The headset is the right sizing/threading for frame and forks.

The local bike shop did not do any work on the forks. I went in afterwards and they seemed to not have the necessary kit to the face the crown race facing on the fork.

Any advice?

Different bike shop?
Go to a proper frame builder?
Fit a cartridge headset instead?

Any ideas to the cause?

The part in bold.
And/or there is something wrong with the forks steerer.

Very often I've seen steel forks behaving that way after a hard front collision, when the frame had no damage, the forks got the worse of it and although could not be spotted, a headset that previously worked without problems would bind right at 90º turn.
Leaning a ruler along the steerer may give some clues.

[BALBAL]

Thanks for all the brilliant comments and advice given; much appreciated.

[fausto99]

as in 'swash-plate' for example.

cheers

So now, after googling, I know all about helicopters! Thanks.