DocRay: <snip> I agree with you that DaveS is wrong about the R3 headset bearing seats, but his assertions are reasonable. Wrong, but reasonable.
After you've read a few posts from cyclists who have damaged frames because the integrated headset was improperly adjusted (too loose), you might change your mind.
Please elaborate on the wrong but reasonable comment, it seems contradictory. Do you really think a thin piece of aluminum is the preferred bearing seat?
Yes, I think that a thin piece of aluminum is a reasonable bearing seat in this application. Cervelo markets the R3 as a high-end, no-holds-barred racing frame. Most owners will have the experience to know that they can't ride on a loose headset, even for a little while. Inevitably, some owners will not know this and will damage their frames. A high-end machine requires high-end maintenance, which I think is reasonable in this case, especially given that weight is a primary design criterion for this product. As you surely know, what's reasonable for any given product is largely a function of the predefined design criteria for that product. In my engineering judgement, the aluminum seat/race is reasonable in this application.
Now, how can I say that your complaint about the bearing seat material is wrong but reasonable, and how is this not a contradiction? Well, I just mentioned engineering judgement. That's an opinion
, no matter who it comes from. I disagree with your opinion that aluminum is inadequate in this application, but I understand why you think so and I think yours is a supportable conclusion. If you suggested that, say, 6/4 Ti was the only valid material from which to build a bicycle, I'd say that you were both wrong and unreasonable, as it's perfectly fine to build a bike from steel, 3/2.5 or aluminum (among many other options).
While I don't mind non-replaceable aluminum bearing seats in my R3, there's a solid argument to be made in favor of replaceable bearing seats made of another material (or the use of a standard headset). You're making that argument. Again, this is something about which reasonable people can disagree. One point on which we do not disagree is that replaceable
aluminum seats would be better.
Now that I think of it, I'm also skeptical about LOOK's use of a carbon fork crown race too.
Do they actually use a carbon race, or a carbon crown race seat?
I understand your skepticism, but Easton has been making forks with carbon crown race seats for years...I use one of their EC90SL forks with such a seat on my own regular-headsetted Litespeed. It works fine. I've never encountered anyone who had a problem with a carbon crown race seat. That's not to say there are no problems, but they are certainly not epidemic.
One problem I've noted is manufacturers have notoriously poor instructions included with frames, forks and headsets (including LOOK). They gloss over the adjustment of the headset, only telling the user to "eliminate play". Integrated headsets need to be adjusted as tightly as possible without inducing excesive drag as the fork is turned. The top cap requires quite a bit more torque on it than would be used with a caged ball headset like Campy's. Failure to do that can result in frame or bearing seat damage.
YES! I totally agree. I suspect people are worried about "overloading" their headsets by applying too much preload, but in fact the danger from having 10% too little preload far exceeds that of having 50% too much preload. Headset bearings are annular and designed to take a large axial load, but people generally adjust them until play is just eliminated, as you say. There's nothing wrong with having a headset that turns a little more stiffly than necessary--you won't feel it while riding and it's cheap insurance against a pitted headset (or worse, in the case of integrated headsets).
I'll agree that if you like the ride and handling of your R3, that's great. I found the geometry in the 49cm and 51cm (which I owned) to be unreasonably short in the front-center with too little steering trail. There is no other brand that makes frames with such a short F-C. The excessiely short chainstays also caused problems with my triple crank (for climbing the Colorado mountains). Unless the shift from the middle ring to little ring was made from the third largest cog (or smaller) the chain would drop off the little ring and jamb between the little ring and the frame. A chainwatcher won't fit onto the downtube either. I was so unimpressed that I sold mine after 200 miles of use.
Yeah, I agree mostly. 54cm is the first size where the dimensions start to look normal to my eye. A 54 has a 54.5-cm top tube, and I'd much prefer a 55 or 55.5 cm top tube. The chainstays are fairly short (40 cm) but that's not really a problem for me. If I rode a 49 or 51, I probably wouldn't buy an R3 either.
Personally, I'm neither surprised nor troubled by the fact that the R3 doesn't deal gracefully with a triple crank. As I said before, it's a no-holds-barred racing frame, and for most racers, a triple is low on their lists. If it were me, I'd probably run a compact crank rather than a triple, especially given the short chainstays.
Finally, you complain about inadequate trail on smaller R3s. They've got 73-degree head tubes (72.5 degrees in the smallest size), which is pretty reasonable. I would assume they're using forks with 43-45mm of rake, which would result in reasonable trail numbers. Some manufacturers use forks with more rake in the smaller sizes to avoid toeclip overlap (an outdated term, but still an issue), but I don't believe Cervelo is doing this. So in the smaller sizes the wheelbase gets pretty short, but I don't see any unreasonably low trail numbers. Longer front center might be a good idea, but I don't see the "too little steering trail" you mentioned.
As an aside, I'm amazed by how many riders in Paris-Roubaix are said to use forks with increased rake "for greater shock absorption." Increasing fork rake decreases trail, which thereby decreases stability. An increase in fork rake will get you a tiny increase in fork deflection (look at the governing equations from beam theory) and a much larger decrease in stability. It's hard to build an unrideable bike; zero- and even negative-trail bikes are perfectly rideable as long as you don't try to ride no-handed. I'm sure these guys go faster because they're more confident in their modified equipment, but objectively, there's nothing there.