the sci guy 

So I know the usual problem is headsets coming loose - but lately I seem to be having the opposite problem. I'll get it where it seems to be nice and snug with no play, but it still turns freely. I can even rotate the spacers underneath the stem with a little effort. Then either just by sitting in the garage, or after a ride, the headset has tightened enough where the steerer snaps into certain pitted positions. And this is a brand spankin' new headset. What the what? This is a regular threadless headset. External cup top and bottom (Cane Creek 40).

I am tightening the top cap before tightening the stem bolts as well. It almost seems like tightening the stem bolts makes the headset tighter, but that definitely shouldn't have any bearing on it?

andrewclaus

Metal steerer/stem? How old is the headset? Has this been happening since installation? Any changes to the bike lately?

Maybe the easiest thing to check is to try a different stem.

Nice little pun in the last sentence, by the way.

the sci guy 

yes alloy steerer and alloy stem. brand new headset and a fresh build (older frame/fork, but all new everything parts). Yes has been happening since i started riding it last week.

WizardOfBoz

As your note points out, the cap should adjust preload, and tightening the stem clamp should not add or reduce that preload. The fact that you can rotate the spacers before you ride indicates that the preload was not affected by stem tightening.

Did you use assembly paste on the stem clamp? If not, the one theory I could come up with was that during your ride the pounding on the stem by your weight on the handlebars and hitting bumps and stuff forced the stem lower on the steerer.

BTW, for a brand-spanking new headset there should not be any detents where the fork settles in. The turning should be as smooth as a baby's bottom.

Good luck and please post any learnings.

On Edit: I see that you have an allow stem and steerer. When you tighten the stem, does the stem fully close, or is there space between the two sides that the tightening bolts pull together? If they are bottoming out, perhaps the stem hole is too big?

Rolla

Add a small (1 - 2 mm) spacer under the stem or under the top cap and see if it solves the problem.

Crankycrank

I've seen where some spacers and stems can have a sloppy fit so when tightening the cap, the headset feels fine but after tightening the stem bolts the stem/spacers/headset will not be perfectly flush and cause weird problems. Sometimes it helps to have your stem cinched up a bit so it can't wobble when tightening the cap for preload but still loose enough that it can move during adjustment, then tighten the clamp bolts.

holytrousers

I would rule it out as the sci guy said "... either just by sitting in the garage, or after a ride. .."
We need more information, pictures, details, etc.

SoSmellyAir

Maybe some debris within the top tube is getting dislodged after a ride and said debris is causing the steerer or bearings to slightly bind against something?

WizardOfBoz

But these are sealed bearings...

delbiker1 

It sounds to me like the top cap/spacers are not square with the top of steer tube. I have had that issue with the spacer and the top cap off just a bit. Centering the spacer, then centering the top cap on the spacer, solved the problem. This may not be your issue, but worth looking at, anyway. I like the idea Crankycrank gave with slightly tightening the stem bolts to assure it is staying in place.

trailangel

Have a pro install the headset. They have the correct tools to face and ream if necessary.
Did you just pound the cups in yourself?

the sci guy 

i tinked the cups into the headtube with a mallet just enough so they'd stay in place, then i swung the frame headtube first into the garage wall as hard as i could to bang it all the way in. was that wrong?

I think we might have a winner here. I made sure to crank on those stem bolts after fixing it again. After ~30 miles this morning, the headset felt just the same as it did when I started the ride. the zipp stem came with star bolts, and I don't have a torque wrench that'll take a bit like that (i only have the handheld preset one thats a standard 4mm hex). Guess I wasn't hand-tightening it enough. Let's see how it goes after another ride.

Thanks for the idea!

WizardOfBoz

Generally it is wrong to hit any bearing surface or assembled bearing. When one does this, it can "Brinell" the bearing races (put little dents in them) and squash bearings a bit. Class 9 bearings have a runout tolerance of 1.3um (millionths of a meter, or thousandths of a millimeter. This is 40 millionths of an inch.

Once installed, the bearings should be very square. In this way, all the bearings take any impact at once and the load is distributed.

One can install the bearing seats using an install tool (a tube with a mating surface so that any whacking one does reinforces the shape rather than distorting it). The best way to seat bearings and bearing seats is to use a press or a proper tool (can be all-thread with some washers, spacers and nuts, btw). and to use this in such a way that PRESSURE AND IMPACTS ARE NEVER EVER TRANSMITTED THROUGH THE BEARINGS DURING INSTALLATION.

So yeah, installing any bearing by swinging the part into the garage wall as hard as you can is ... highly non-standard practice.

But if the bike rides nice now, you probably are ok.

Outrider1

It depends on the garage. If left sitting, the headset will either tighten (as in your case), or loosen (as in most cases).

trailangel

[QUOTE=the sci guy;22549648]i tinked the cups into the headtube with a mallet just enough so they'd stay in place, then i swung the frame headtube first into the garage wall as hard as i could to bang it all the way in. was that wrong?
/QUOTE]

LOL.... if you had concrete walls in the garage.... probably will be ok... but just drywall will not seat the cups.

the sci guy 

I appreciate the time and thought you put into this, but I was definitely just kidding. The headset was installed with that $200 park tool headset press. It's all nice and flush and proper.

yeah but i hit it right where there was a stud behind the drywall!

SoSmellyAir

You need this (5 Nm): Topeak Nano TorqBar 5 with 5 Tool bits, 5nm : Sports & Outdoors (amazon.com)
Or this (6 Nm) version: Amazon.com : Topeak Nano TorqBar 6 with 5 Tool bits, 6nm : Sports & Outdoors

Each comes with two Torx bits: a T20 and a T25.

WizardOfBoz

Wow, I use a piece of 3/8" all-thread. I'm jealous.

Not a problem to write this up - perhaps it will cause others to not use the "impact method" of bearing seating.

SoSmellyAir

Maybe I am missing something, but aren't we talking about external headset cups? If so, you just press in the cup, and then drop the sealed bearing cartridges afterward, so there is no risk of damaging the bearings themselves, right?

WizardOfBoz

I think that some terminology should be clarified. Cups and cones (to my mind) are precision-ground parts that actually interface with the bearings themselves. A bearing seat is something into which you insert a sealed bearing assembly. Typically, in the old days, you pressed a bottom cup and a top cone into the headset, and a cone onto the top of the fork. The you'd have a threaded cup that screwed down onto the bearing assembly (balls plus retainer) on the top cone. Generally you'd need some force to mount the top cone and the bottom cup into the frame. For ensuring those parts were seated, you'd use an expensive (relatively, for one use) but durable and very user-friendly Park tool, or some all-thread and some washers and nuts. I don't think it's a simple "drop in" into the frame. At least it wasn't for the old Campy headset bearings. You needed a press, or a mounting tool, or the allthread.

Bearing seats, as opposed to cups and cones, are the things you drop sealed bearings into. These too require some force to ensure proper seating. At least that's what I call 'em.

the sci guy 

what about my rods and cones?