I'm waiting for some images of this past weekends riding - some mildly epic paved and moderate unpaved. But my mind is wondering this morning to the techical.
Front end shimmy is an issue that rando bike builders must deal with. As a basic rule, it seems that front end loads coupled with light tube sets can lead to the problem. Interestingly, a change of headset can make a radical difference.
In the latest issue of Bicycle Quarterly, a Toei was reviewed, and it had a real shimmy problem. I can't imagine that it is the bikes fault. The tube set is not terribly light, and they've been working on geometry for decades. But unlike most bikes in Japan I suspect along with the reviewer, that the "trouble maker" on the bike the Chris King headset. Now the Chris King headset should not be considered a "bad" headset. If anyting, it might be too good! Let me explain.
(Now I'm not sure I have this all straight, but here is my understanding of this topic)
If you spin a wheel in your hand, it wants to "wander" - it goes into a bit of side to side ocillation. Now when the wheel is on the bike, and your moving along, it does the same thing and it rotates the handlebars along with it. Now a bike frame has a natural frequency for its own movement (others can explain this better). If the wheel's frequency of ocillation is in sync with the frame's frequency, together they build, and a shimmy in the frame develops. And all this is related to frame geometry and forkrake as well. I'm sure glosses over something, but it captures the core idea.
For a frame that has a shimmy, anything that is done to change the frequency of either the wheel/fork or the frame, to get them out of sync, can reduce the problem. So a much heavier or much lighter frame can fix things - so a frame may even be "not light enough" to avoid a shimmy. Of course, going super light can have other problems.
One easy fix to imagine is to change the dampening of the headset. A Chris King headset rotates very easily, so with many tube sets and geometries that are otherwise favorable to randonneur riders, it seems to have a tendancy to get systems in sync so they shimmy.
But why to needlebearing headsets, such as the Stronglight A9, or the Miche shown above seem to reduce the problem? Well these headsets have natural dampning! A wise friend who hadn't thought too much about this opened my eyes to the probable reason. With a needlebearing headset, especially one that is not made with tappered rollers, the needles are only in true rotation with the races at one spot on the needle. This is because the race has variable diameter - so where the needle is contacting the race part with a larger diameter, the needle would need to be rotating faster than along the area with the smaller diameter. So with the needlebearing headsets, the actual rotation contact area is small, and there is "sliding" along nearly all the remainder of the surface! On the plus side, there is an increadible amount of contact area, so the headset itself wears out very slowly, but it doesn't move as freely. So a needlbearing headset is a headset with natural dampening.
Now one could even play with grease viscosity to fine tune the system. Of course, the handling of the bike then becomes susceptible to temperature change, but this happens anyway.
Now of course, some bikes might "avoid a shimmy" if dampeining is bad (casuses a frequency match-up). So maybe some race bike shimmy less if a Chris King headset is used. Interesting question. But for us on the rando side, the needlebearing headset is a help.
Is there a downside to the needlebearing headset? For one, I think that at low speeds I can sometimes feel a headset with a lot of dampening as it is harder to correct in sharp turns. But I'm not quite sure about that. Certainly, with hands on the bars, using a needlebearing headset is rather impossible to detect.
Now I have some more thoughts on this too - but those will follow in a few days. It would be great to get some comments on this!