I'm open to suggestions on this one. Cantilever front brakes, Kool-stop black shoes. Under front wheel braking there is a rhythmic <thunk> — one per wheel revolution. This started suddenly going down a steep hill. I figured a front ball bearing had gone and replaced them. But the balls and races and cones all looked good. And the <thunk> is still there.

When you look down on the fork you can see that it is being forced backward briefly, like the brake had suddenly gone to full lock then releases within ~5° of revolution. I can't feel a pulse in the brake lever but it's hard to tell since the <thunk> shakes the whole frame. Intensity is proportional to load. Can't feel it on the stand. Spun the wheel: the rim and tire look fine. I don't see anything like a bulge or dent on the brake surface.

I suspect the <thunk> itself is a result of too much play in the headset, but the cause has got to be in the wheel and brake interaction.

Ideas?

Since you say the rim looks fine (nothing obvious), I'd guess it's the rim seam combined with a loose headset. It might not be that big a deal otherwise, but with a loose headset it wouldn't take much of an inrregularity in the rim to cause noticeable movement of the fork while braking, and on inexpensive rims particularly, you'll rarely have a smooth rim seam-

Sounds like a seam or blip to me. I have a Mavic x-517 rim that does this. I'm not sure if that is an inexpensive rim, but it is the one that came attached to my XTR hub.

Stand next to the bike and push it slowly while holding the brakes on slightly. Can you feel the drag? What portion of the rim does it match up to? I suspect it's the seam. Or if it's some other section, pull out calipers and measure the width and make sure it's the same all the way around.

Quite inexpensive rim - cheepo bike.

The headset doesn't feel loose when you grab the fork and try to move it, but I haven't adjusted it (threaded) yet, nor have I been into the bearings.

I couldn't feel a seam with my fingers.

Either the rim seam or the rim is out of true.

No.

Haven't identified it yet. Since I can't make it happen on the stand, I'll put some chalk marks on the tire and try to ID it on the road.

I have a micrometer that might clear the tire. That would be a lot easier than calipers. I couldn't feel or see anything, though.

I did find one spoke (of 36?) kinda loose.

I'm surprised you'd be able to feel the seam on the brake surface of an X517, those seams are welded and ground smooth at the factory, not pinned seams. I've got two wheelsets with X517's, if not for the visible seam on the inner portion of the rim, I'd not be able to tell a seam is there at all. I'm not doubting you, I'm just saying it's surprising it would be noticeable on those particular rims-

Is one pad contacting the rim before the other and "pushing" the rim out of true under braking?

Nope.

Rim is true, too.

What about surface-finish? Is the rim uniformly smooth all the way around? Also overheated pads can sometimes deposit a thin invisible layer at certain parts of the rim that's not visible. Cleaning the rim with lacquer-thinner will remove old sew-up glue and brake-pad deposits.

To check headset play another way, stand over the top tube, apply the front brake hard and rock the bike forwards and backwards without lifting the wheels. If you feel some thunking, you can and should tighten up your headset (after making sure bar-stem, stem-fork, and fork-wheel interfaces are tight). Then you can check the thunking with regular wheel braking while the wheel is going around.

Well, I finally had occasion to investigate this further. Spinning the wheel sans tire there is no apparent deformation or out of round. Sidewall is way too thick to be flexing. Mic'ing across the out edge of the channel shows a consistent 1.0750" ± maybe .005. But at one very short - 1.2" or so - section it flares to 1.0870". I can't help but put this down to damage, but there is no indentation or other evidence of impact, and I can recall no incident in the approximate time frame. It must have been recent because there is no evidence of additional brake wear at that point. Is a puzzlement.

I'm also surprised that a mere 12 thousandths would cause such a dramatic symptom...

I think I'll try to reduce the flare with a bench vise. Or I could try a hammer and dolly. Any other suggestions? Should I use heat?

Have you tried reversing the wheel so it spins the other way? Your rim's defect might rub the brake pads different going the other direction, and if you put the wheel in "backwards" your problem might go away. I've done this on a rim that had an uneven seam.

That hump that flares out would easily produce the sort of pulsation you're feeling. Often such a thing occurs at the joint of the rim if the weld wasn't dressed correctly.

If this were me and I wanted to save the rim and if the rest of the rim was quite consistent I'd mark off the extent of the flare's edges and peak on the inner spoke arc of the rim so it can't be scuffed away and then I'd carefully work down the excess metal with a fine cut file. I'd also use a good straight edge to see if the flare is all on one side or if it's even on both sides. All of this fine measuring would determine how I proceeded with dressing away the excess metal.

That's more than enough to cause the pulsing. I'll usually notice width-variations larger than 0.1mm or about 4-thousandths. A lot of it also depends upon your pad-compound. Hard-pads tend to make it more noticable while softer pads are more forgiving.

Kool-stop black (I think) on cantis.

Some good ideas here. Thanks.

Well, I spent some more time with the wheel. This is a curious injury: there is a "dent" in the rims on both sides, as one would expect were the damage caused by running over something (though one would also expect a pinch flat), but in this case the dents are in the same direction - 'in' on one side and 'out' on the other! - so that the spread between the two rims remains nearly constant. (The long pads never 'notice' the innie - only the outie.)

My talented imagination has come up with no scenario that could cause this. :confused:

Anyway, the hammer wasn't the answer. The mill file was. Thanks BCRider.


FastJake, I didn't try reversing the wheel (probably won't have worked in this case), but I'll keep the trick in mind.