Techie Tuesday: Pt II - Headset R&R on Specialized '06 Roubaix Elite

Note: This is part II of a 2 part series. Part I can be found here.

The Cane Creek 110 IS41 stainless steel headset arrived last week, and I set about installing it, but it didn't fit, at least not until I got real creative, and reused the OEM crown race as a top bearing spacer. Before I forget, it looks like Universal refunded my return, so kudos for them.


Red and black and rock solid for 110 years. Note the very tiny gap between the headtube face and the bearing cap. The hint of blue between is the bearing cap's rubber seal, and seals the headset bearings from the weather.

I hope Google finds this pair of posts noteworthy, and thousands of owners with AheadSet, or MindSet, or whatever headsets, can do this without the gut-wrenching fear of destroying their frames I endured.

My gut tells me it will be mostly professional mechanics that end up using these two posts, and that is why I took the trouble to take so many photos. Many of the details will not mean much unless you're neck deep in fear and uncertainty, and then they'll be priceless.

If your bike is a year old or more, and you have one of these non-standard headsets, I would strongly recommend you replace it with a standard Cane Creek, system (Cane Creek AKA Dia Comp, invented the Integrated System ~ 20 yrs ago).

You can go top of the line with the 110, like I did, and get a 110 yr guarantee, but the 40-series I started with is also perfectly acceptable, and may well last the life of the bike for those putting 2-3,000 miles per year on their bikes.

An important physical difference between the 40 and 110 bearings, is the entire depth of the 110 bearing is full diameter, whereas the 40 bearings are smaller in diameter after 3-4mms. The 110 bearings are therefore much less likely to ovalize in the headset.

The longer you wait, the greater the chance that parts and mechanics are going to be unavailable to fix your headset. (assuming you can't get a standard replacement headset to work, or you need some OEM part to make the new system work properly)

My go-to mechanic, Eddy at MadCat, a very wrench-centric shop, had only seen one other headset like mine, and that was on a mtb. Time will eventually paint you into a corner, and force you to abandon your frame. Don't let that happen.

------

As soon as I got the bike home from the shop, I cleaned the headtube thoroughly with 99% rubbing alcohol, and studied the old and new systems to make sure I understood how they worked, and fit together. I also inspected the 45 degree flanges milled into the aluminum., bonded sub-structures.

When I got the new CC 110 headset, it was immediately obvious that the crown race was not going to fit my carbon fork. I was pretty disgusted that the bottom of the steering tube, right above the crown-race bulged area, wasn't even fully epoxied. Some of the carbon weave was exposed. That's one of the most critical areas in the fork, so that's an inexcusable lapse in QA, and a serious safety issue. Nevertheless, I had to find a way to reduce the diameter from 30.10 to 30.00 mm.

I bought a set of assorted 3M Wet Or Dry sandpaper at the autoparts store, cut a 4x8" sheet of 320 grit into 8mm strips 8" long, wet them, and worked my way around the fork, turning it about 30 degrees on each turn, while sawing the sandpaper back and forth to engage 3 sides at once. I paid special attention to a very rough area that looked like it had been cut with a crown race tool meant for metal steering tubes. Another serious flaw in the OEM fork.

After 2-3 times around the steering tube base, stopping each time to slip the new CC crown race down to check the fit, I was getting pretty close. The crown race would rock back and forth over the high points, so I made marks on the steering tube with a grease pencil and focused on those. (99% alcohol removes the grease pencil markings)

I was able to get the crown race down to within ~ 3mm by pulling it down with my finger tips, and knew I was getting close. You need a very snug fit, but can't risk bending the race, as it is very thin on the CC system.

About that time I noticed the very bottom of the steering tube, where the base of the race would engage it, was a bit wider, and tapered slightly. Sandpaper just doesn't have enough structure to make a nice, clean 90 degree interface between the top of the fork and the bottom of the steering tube.

After some head-scratching, I decided to use a high quality, diamond impregnated finger-nail file, and carefully work my way around the base of the tube. That worked very nicely, and after a final light sanding, I did one last trial fit, and decided I was close enough that I could get the crown race to seat very snugly, and all the way down - although you never really know until you try it.

If you have an extra $13, I would suggest ordering an extra crown race with the headset. It's always a lifesaver to have an extra, and I wouldn't have wanted to have waited another week for parts.

As it was, even without grease, the crown race seated perfectly, and the "machining" of the steering tube with 320 grit wet or dry, left a very smooth finish. I never did use the 400, 600 or 800 grit included in the assorted package.

You never want to take too much off the steering tube, and have a loose fit. There's no good way to recover from that. You've just ruined the fork. In that situation you could try using some silicon caulk, but really, you've ruined the fork. With the crown race in place, I trial fit the bearing in the headtube, and checked the seal. PERFECTION!

The bearing facing looks pretty rough, like a file was used on it, but it doesn't affect the function. I am going to send this pic to Cane Creek though and ask them if it's been used or this is factory spec.

Topside assembly with OEM crown race retasked as a spacer



Side view of topside assembly with the OEM crown race showing under the top bearing. Note that the 110 series bearings are the full diameter for their entire depth. The 40 series are not.

45 degree bevel on OEM crown race, and 30mm ID make it a perfect spacer, and saved me from having to reface 5-6mm off the top face. Smooth as glass when installed.

When compressing, align the two splits

Split realigned under compression when fully assembled by rotating the top cap 180 degrees

Even while trial fitting the bottom bearing, it was clear the top bearing was going to seat far too low to keep the lip seal on the bearing top cap from betting smashed against the top face of the headtube - and by 5-6mm. My heart sank. Was this even going to work?

After greasing the top of the steering tube, I slid the top cover down, careful to not damage the internal O-ring seal, and confirmed my suspicion. (CC's  O-ring seal is VERY tight. Grease and then wipe dry with alcohol before clamping the stem down)

One option was to have the top of the headtube refaced. I check, and the Part Tool facing tool was $450. It was 10:00 at night, and I really didn't want to drag the bike down to a shop and hope they knew what they were doing. I also had no good way to know now much I needed to machine the headtube down, and it should be fit to less than a half a millimeter, as the shims are 0.25mm each, and there is only room for 2-3 on top of the compression ring.

Frantic, I looked around for anything I had in my parts bin that might work, and then looked in the project tray I'd put the old headset parts in. The OEM split crown race looked promising. It SHOULD be the right dimensions, the right angles, and hopefully, enough material to take up the extra space. It worked!

As you can see in the pics, it's pretty beat up, and I have ordered an FSA Orbit replacement crown race, but flipping that crown race upside down and placing it over the compression ring works great. It might even add some strength to the headset in the process.

I initially aligned the split gaps on the compression ring and OEM race, so they would compress together, and not gouge each other, pushing their edges together when compressing. Once on the bike, I used my hand's thumbweb area to rotate the loosely compressed headset cover/bearing cap 180 degrees. When I pulled the headset apart after a short 20 mi ride, they were still on opposite sides, as you can see in the photo. This is probably the strongest, and best centered arrangement, so I recommend it.

I tightened the compression cap down hard to really compress the stack, making sure the crown race, bearings, compression ring, and OEM race were all fully seated. Having a bunch of extra spacers meant I could do a lot of the trial fitting without messing with the stem and bars. Nice tip. Unfortunately, the lip seal was too tight against the top headtube face, so I needed to figure out how to use the 0.25mm shims I'd bought.

Taking the whole assembly apart again, I "peeled" the compression ring out of it's captive groove, machined into the bearing cap assembly, added 2 stainless shims, and reinstalled the compression ring. Reassembled. Too much. I could see daylight between the face and the seal. Disassemble, remove compression ring, remove 1 shim, replace compression ring, reassemble. PERFECTION!

I used a small pair of side-cutters/wire dykes to remove the compression ring, grasping the edge of the split, but did leave a small engraving on it, so a high quality pliers would have been better. A nose pliers will not work. Too flexible and not enough leverage. I didn't make a video of that, but maybe I'll remember when the new FSA Orbit crown race arrives and I have to pull the whole top end apart again.

Zero stack height, InterLok to flat surface, brass ring adapter. It allows Cane Ceek 110 series headsets to interface directly with the bottom of any standard stem. My personal RX would be to buy some insurance and top that off with a 0.25mm or 0.50 stainless steel shim/spacer. These shims are meant to be used inside the bearing cap to raise it enough that the rubber lip seal just touches the headtube face - without any daylight showing between them.

CC brass interface spacer in place in an InterLok nylon spacer

This simulates the top of the bearing cap on a 110 system, brass ring and stainless shim in place, ready to interface with bottom of handlebar stem. An intriguing idea is to use this setup, but replace the brass ring with an O-ring to get a primary weather seal right at the bottom of the stem's cut, which will leak water like a sieve in the rain.

Nylon spacer, stainless steel shim, and brass interface ring.

-------


Having been through all of this, I have to say, while integrated headsets are a clean solution, Chris King has made some very good points as to the folly that motivated this design. External bearing cups are just better. Period. They ensure that whatever happens to a headset, the frame is never in jeopardy.

Of all the integrated designs, I think the ZeroSet is the best, as the top of the bearing cap also comprises the sealing surface on top of the headtube. It means you won't need to reface your headtube to get a perfectly smooth and flat sealing surface. My frame, unfortunately, needs some smoothing out, as I'm sure, most IS style frames do over time.

Cane Creek ZeroStack -  AKA ZeroSet. The press-in steel insert extends over the top of the headtube's face, adding strength and a perfect lip-seal mating surface

I'm also going to write Ceradyne ceramics, and suggest they manufacture the inserts that receive the sealed bearings, which are bonded to the carbon headtube, out of ceramic. They make almost all of the body armor used by the DOD, all of the helicopter, and most of the vehicle armor as well.

Chris King's very real concern, is the back and forth rocking of a headset can ovalize these inserts, commonly made of aluminum, until the sealed bearings start to rock back and forth in the frame - rendering it useless.

Titanium diboride ceramic is about the same weight as aluminum, but is tougher than tungsten, and about any other substance on Earth. It might cost a frame maker $50-$100, but would make integrated headsets light, and bombproof.

My OEM press-in bearings used a deep penetrating insert, which fit into, and past, the 45 degree bearing support flange to prevent ovalizing of the aluminum bonded sub-structures. That was a good decision. Having no effective seals, unforgivable. I never rode in the rain because I knew the headset wasn't sealed, and look how utterly destroyed the headset was anyway.

I'm a lot less impressed with Specialized as a brand, because they are doing far too many things, in headsets, and bottom brackets, that make their bikes hard to service - IE: disposable. It's also worth pointing out that upright "comfort bikes", with long steering tubes, and lots of spacers, place a lot of extra strain on the headtube, and headset bearings.

Asked for a RX for a new $2,500 bike last week, I recommended a Motobecane mail order special. Until you get above $4k, I just don't think the name brands are buying you much, and the prices are almost double. A name brand should buy you dependable, reliable QA, and so far, I haven't seen that.
.

The Thrill of Victory

I headed out the door early yesterday, about 4:00pm, with the heat of the day near its peak, but as it was a cool day, it was quite comfortable, although I was sucking on my waterbottle within a mile, as it was also dry.

I had intended to ride my favorite 65 miler to Rescue the day before, so had changed out the 12-23 cassette out for the 12-27 I usually ride. It was nice to find old friends on the shifters again! 12-23 is great for TTs, as tired legs need an optimal cadence, but with the small climbing rings on the triple up front, it makes for gears that are tediously granular for hills. Too much shifting to find the right gears.

It always surprises me how different things look in the bright, overhead light of mid-day, vs near dawn or sunset - especially along the American River. I had a chance to appreciate the different view, and the dense shade in places where the wet winter and late spring have the trail very overgrown - especially with berry briers.

Trail repair is underway just past Hazel, with some retaining wall work being done. The detour route I quite like, and will continue using it even after construction is complete. It is more sheltered from the headwinds than Gold Country, and not nearly as congested as the bike trail from Sunrise to Hazel. As for the construction on the Hazel Ave Bridge, it is still underway, and the detours there were compounded by a malfunctioning crossing light.

After waiting for 3 light cycles, I risked it and headed across with the auto's green light. Since they all turn left, if they don't stop, you're going to get hit. After 3 cycles though, there were 12-15 riders waiting to cross, so the auto traffic graciously waited before continuing through the intersection.

I was being tail-gunned by a rider from the Aquatic Center until well-past Blue Ravine. Eventually I waved him around and returned the favor. He was trying to drop me, but there was no chance of that. Not strong enough. It did get my competitive juices flowing though.

I stopped for water across the bridge, winding my way down the sidewalk before stopping at the little outhouse. Whew! God that thing needs cleaning. I held my nose and refilled my waterbottle, making for some pretty weak Gatorade, but it was hot enough I needed the fluids.

I wasn't feeling all that strong pushing off, but was happy to have the stench behind me, so I hit the Garmin lap button more out of habit than anything. Once past the footbridge though I caught a tailwind and it kind of got me going. I hit the base of the first climb in good form and pushed myself. By the time I went under the Folsom Crossing Bridge I was winded, so backed off a bit on the straight stretch that runs parallel to the road there.

Not a PB, but winning a race is always thrilling!

About 100 yards before the turn that starts the 2nd climb, I got passed, and decided to try to stay with him. Good motivation to push myself, if nothing else. He was riding the same bike as I was, and I think this is the same rider I raced a couple of months ago coming home from Hagen Pk.

He tried to drop me for a half mile, and then settled in, I assumed to wait for the steep pitch at the top. A good move, as I outweighed him by 30 lbs, so he was picking his field of battle well. I focused on good body position, sliding forward on my seat and riding upright on the hoods, using my glutes as much as possible, and breathing deep to get my HR down.

Sure enough, 250 yrds from the top, where the steepest part of the climb starts, he stood and started to hammer. I stayed in the saddle, gripped the very front of the blocks, slid forward just a bit more, and dug deep. I'm not sure who was more surprised, him or me, but I went right past him!  >B 

I kept expecting him to put on a move and pass me, so I kept the hammer down. My pulse rate was skyrocketing, but we only had a hundred yards left at that point. My quads were screaming, but no way I wanted to lose now, so I kept digging deeper.

In the end I beat him by ~ 25 ft. Damned, that felt GREAT!

After a nice rest and conversation up at the Beals beach area, I headed back down for home. Just past Negro Bar (originally the campsite of gold miner's slaves during the California Gold Rush) a green Colnago blew past me and kept up a blistering pace until out of sight. He did seem to weaken a bit on a short 6% climb though, so I kept after it and made it a challenge to see if I could reel him in - eventually, maybe.

I caught him ~ 5mi later climbing up the new Hazel Bridge approach. Apparently my attention to aerodynamics and likely greater power was enough to close the gap. He would have killed me on any climb, but on the flats, headed into a 10-15mph headwind, I had the advantage. Moving my waterbottle down 3" has made a bigger improvement than lowering my aerobar position.

With some nice pick-up racing thrown in, I arrived home all smiles. Another fun day on the ARPT. I can tell I pushed myself to 103% of max HR today, but well worth it!

.

Dead Weels A'rollin

I went out the door for a quick ride Thursday, noticing at the last moment how little time I had left before sundown. Intending to head off to Beal's Point, my bike kind of went right instead of left at the Sunrise 'Y' , so I ended up heading downstream to William Pond Park. It was a nice day with a 5-7 mph headwind and I picked up a trailer somewhere around Hagen Park.

I felt slow and sluggish, so when he passed me I let him go a bit and then realized he'd pretty much spent himself getting around me, so I bridged up and rode his wheel for awhile. We ended up riding up over the WBP bridge side by side carrying on a nice conversation.

I was feeling a lot better by then - all the rust was falling off and I had some snap in my legs - so I decided to head downstream to CSUS. Doug bailed about a mile down from WBP, but I put in a plug for HammerinWheels and hope to see him again on a club ride.

I picked up the pace, hunkered down in the aeorbars, and tried hard to slice through the the wind while keeping the sweat from dripping off the tip of my helmet and spattering all over my glasses. I was blowing hard as I rode up the long ramp to the Guy West Bridge at CSUS. Amazing how much of a lather I can work up in 60 degree weather at 20 mph.

Eager to trade in the headwind for a tailwind, I hit the Garmin's lap button and pushed off 5 minutes later. I set a pretty solid 22-23mph pace and passed a half-dozen riders along the way. As usual, I got no challenges along this section, so I had to focus on my Garmin and kept pushing myself till my HR monitor would alarm and then back off after a minute or two.

After passing a fast commuter with rear panniers filled, I put on a sprint to keep him off my wheel. No point pulling the kitchen sink. Just as my HR alarm settled back down I ran up onto the back of the mythical Park Ranger SUV (it exists, really, I swear!) doing a very proper 15 mph - the legal speed limit on the ARPT.

If anybody actually observed that speed limit the bike trail wouldn't be much use to anyone, but there he was, stuck in my face anyway. I should have peeled off the trail for surface streets at the edge of the WBP park, or at least hit the lap button, but as it was I ended up crawling along behind him and ruined my average.

Once home I found I had spent about 50% of the ride in Zone 5 - low Zone 5, fortunately. I often ride the hardest going downhill or with the wind at my back. Speed begets speed.

Last night I was cleaning and prepping my bike for a ride tomorrow from Auburn to Foresthill - a nice climbing route - and decided to finally Scotchbrite my rear wheel's braking surface clean to get the undulation out of the rear brake. I was shocked to discover a spoke breaking out through the rim, splitting it in the process.

This is not a welcome expense, with winter clothes and Christmas to budget for, but at least I didn't end up sprawled all over the road coming down a 10% grade at 50 mph. Besides, I have a zero-guilt reason now to finally put some good wheels on the bike. They have always been the weak link on my Roubaix.
.

Are You Wasting Power?

Women's Ruby with SRAM Red


Specialized Roubaix with SRAM Red. Note how top of sifters are ~ level with the bar top. This provides the maximum amount of forward reach for climbing


Tarmac racing bike with SRAM shifters and racing handlebars

Looking at the number and size of spacers, do you notice a trend here? Look at where the headtube's bearing race is relative to the top of the top tube on these bikes. With the arched toptube of the Tarmac, the bearing race is actually BELOW the top of the toptube - not sticking up 25mm (1 inch) above it like on the Ruby.

Look at where the top of the handlebar drops are - your drop position - relative to the bottom of the top tube. Now fire up the bike calculator (upper left of the blog) and remember that with 25mm of extra headtube height, a 20mm bearing collar, plus a 10mm and 2 X 5mm spacers, even when riding on the hoods your Ruby's position is "bar tops" for this calculator.

For a 150lb rider putting out 200 watts into a 5mph headwind the difference between "bar tops", and drops that are actually drops, is ~ 3.5 mph, and relative to aerobars, is almost 4.5 mph. Oh, and those aero bars are only actually aerodynamic IF they are mounted on handlebars mounted at a reasonable height. Now enter a speed of 20.07 mph in the Power from Speed calculator and see how much power you are wasting.

To go the same speed as your 200 watts will power you to in properly mounted aerobars, you'd have to generate 349 watts riding on "bar tops" - which is equivalent to riding on the hoods with a big stack of spacers under your stem. That's a 75% increase in required power to produce the same speed.

The best handlebar height is the lowest one your low back will tolerate. That will almost always change in your favor over time as your body adapts to the new position/s. You don't have to go all the way on the first date, you can move spacers from below to on top of the stem over time, but usually it's best to be aggressive as most people underestimate their body's ability to adapt. Ergonomic changes on a bicycle are a convergence process. Body adapts to machine, and then machine to body, and so on and so forth in an endless cycle.

The single most important factor in how comfortable you are on the bike is how much weight your feet are bearing, and thereby, how little your hands and butt are bearing. How will you ever get strong legs riding in a weak, upright position? Marketing departments like to sell plush bikes with seductively comfortable positions to new riders, but any serious rider should look to change such a stock setup ASAP.

So why set up bikes in such an upright position? Maybe so they can solve your speed "problem" with a more expensive bike. I'm guessing. I don't really know. I just know that 4-5mph is a HUGE increase in speed, more than going from a $500 bike to a $10,000 bike, so get out your 5mm hex wrench, maybe your hack saw, and enjoy the fastest ride of your life!

Muzzle me NOT

One of my pet peeves is responding to a product review and then finding out that since I offered actual constructive criticism, instead of an unrestrained sunshine butt blow, it gets tossed out by the site management. This just happened to me here. Then I remembered, I now have an independent voice that no one can muzzle. Without further adieu, here are my comments on this glee factory of a review.

Generally a nice review, with a few caveats from a guy who owns a Roubaix.

First, the 50/34 chainrings make ugly gears with collisions at (34-14, 50-22) and (34-16, 50-25) so two of those gears duplicate each other, meaning you only have 18 unique gears. You need the 34 chainring for climbing, so pitch the 50T and replace it with a 48T TA Alize chainring from Harris Cyclery and get lower and more granular gears.


Ugly gears with collisions and badly placed 34/12 gear

Second, this analysis is referencing a 12-27 cassette, not the stock 11-28, because as you correctly pointed out, the 50-11 combo is a waste. You'll do better just getting in a tuck. The 28T in the back results in a minuscule 0.1 mph reduction in the lowest gear's climbing speed, will make shifting more problematic, and require more chain. Swap it out for a standard 12-27 to get more granular gears in a more useful range. For a guy's bike I can see big chainrings. Guys think about gearing with their little head (how big is it?), not their big one, but gals should be able to stick to the facts and make an intelligent decision here. Specialized should give women more credit.


Beautiful, granular 48-34 / 12-27 compact gearing where the 34/12 gear plugs the 48 / 16-18 cassette gap perfectly

Third, the first time you do a long ride into a stiff headwind you will start looking at tossing all of those spacers under the stem so you can get into a lower position. Ditto for leaning down enough to get a forward position you can climb in efficiently - where your weight is on the balls of your feet. Riding in the flats faster than ~ 15mph you'll have the same requirements. When you lower the stem by removing all those spacers from the stack, the stock stem will be too short.

People who are serious enough about biking to drop $3k on a bike are going to want a reasonable riding position pretty quickly. The stock "plush" setup does them a disservice because getting there will require a new stem and handlebars with a reasonable amount of drop and reach. You will also need to cut the steering tube when you toss 15-30mm of spacers - or ride with a dorky looking and dangerous amount of steering tube sticking out of the top of the stem.

Finally, those tires will start to get very annoying because they have an exaggerated cap on them that will require you to let the air out of the tires to get the wheels past the brakes when taking them off. They aren't horrible tires, but there are many, many better tires that have better shapes, grip and less rolling resistance. Michelin Pro-3 Race are a good choice. All in all, the Ruby is a great bike, but Specialized could equip it a little better.

An added note. You can lower your handlebars by removing spacers or by flipping your stem over so it angles down instead of up, or you can do both to lower it a lot. Stems come in different angles, so you can get pretty low if you wanted to. Generally though, if you are choosing between removing spacers under the stem or flipping the stem, always choose to remove spacers.

The reason for this is spacers move the attachment point of the stem up on the steering tube, and this makes the handlebars pretty rubbery relative to right down on the bearing or with a small vibration absorbing 2-3mm spacer under it. Even small, light riders will notice a difference when climbing out of the saddle. The more upper body strength you have the more this is an issue.


Before cutting the steering tube. Tailor made to break ribs!


U-G-L-Y and dangerous


After cutting the steering tube. Stronger, climbs better, cheats the wind

The handlebar stem bears more torque than any other part of the bike, save the bottom bracket. This is because when climbing out of the saddle and hammering, all of the off-axis torque your feet generate pressing down on pedals - far from the centerline of the downtube - must be offset by torque you generate with your upperbody which is then transferred through the handlebar stem and downtube.

It should also be obvious from this how inefficient it is to climb out of the saddle and waste all of that upper body strength - all of which must be fueled by starving the legs for carbs and precious oxygen. Even the world's best cyclists, when climbing Mt Palomar during the Amgen Tour of California, did NOT stand and climb out of the saddle. If you recall, Lance took the peleton up that very fast to suck the life out of any breakaway attempts - and he did NOT stand. There are times you should stand and hammer, but do so sparingly. Ride smart. Use your energy wisely.

What to Buy?

I was reading an old Bicycling Magazine last night, the annual buyer's guide from last year. I looked at the cover to find the month of issue - April. I resolved to get busy and write something in my bicycle blog about purchasing a new bike. Better me, having just gone through the process 10 months ago, than an industry rag hawking the wares whose advertisers pay the most.

First and foremost you must buy the frame. I'm not suggesting you have to literally buy a bare frame and build the bike up using all hand-selected parts, although that is exactly what I will do next time, but be aware that you can replace every other part on the bike except the frame, so you must consider the frame first and foremost. A good frame is like a savings account, it will continue to pay dividends over time as you upgrade worn out parts and slowly grow into the bike as your conditioning and preferences in rides change.

A good frame should do a lot of things well and not be overly prone to damage, but above all, IT MUST BE THE CORRECT SIZE. Moulton's system is the best one I have ever found. It correctly explained why at 5'9" I should be riding a size 54 frame, and not a size 48 or 49 based on my inseam. I have the feet and torso of a guy 6', and the inseam of someone 5'5". The frame and stem are fit to your torso in this case, not your inseam. I therefore ride a 54mm frame and a 120mm stem with my seat slid forward. It works beautifully. If you are riding a bike with a stem shorter than 90mm or longer than 130mm, you almost certainly have the wrong frame size. Please study this chart thoroughly before you venture into a LBS. The difference between ignorant and arrogant bike salesmen is often 15 minutes.

The 2nd biggest decision in buying a bike with a good frame is the material the frame is made out of. The choices are really steel, aluminum, titanium and carbon fiber. Steel is the classic frame material, and the material has one huge advantage, it has a fatigue limit. Aluminum does not. All aluminum components, including, and especially the frame, will fatigue to zero strength after some number of fatigue cycles. Steel is also the strongest material by volume, and unfortunately for bike frames, it is too strong.

This because as you make frame tubing larger and larger in diameter the strength of any such tube increases with the cube of the diameter. Two inch tubing isn't twice as strong as 1" tubing, its 8 times as strong. It didn't take tubing mfgs long to figure out they could double the tubing diameter, make the tube walls half as thick, and still end up with a tube 4 times as strong. If you continue with this process long enough you will make the tubing walls so thin you can crush them with your bare hands. This is called the "beer can" effect, and forces steel tubes to be made with walls a lot thicker than is required for the strength of the frame. However, steel will only fatigue down to about 80-60% of its original strength and then fatigue no more no matter how many fatigue cycles it is subjected to - we are talking billions of cycles here. The Eiffel Tower, if properly maintained, could easily last 10,000 years, not the month it was designed for.

So, why not use a material that is less dense, like aluminum? You can make large diameter tubes that have thick walls that won't "beer can" and they will still be light. Yup, and they will fatigue to failure in short order leaving the rider in a bloody mess in the street, and eventually, sobbing in front of a sympathetic jury. So why not add in some extra strength reserve so by the time the fatigue factor is getting critical the bike has been melted down into a TV dinner tray?

Well, this is what is done, but the result is very stiff tubing and very stiff frames, and unfortunately, aluminum transmits road vibration like a super-conductor. These frames will eventually fail, but not for many years of normal use. In addition aluminum doesn't rust, in part because only metals with iron in them rust. This does NOT mean aluminum doesn't corrode more or less in the same way as steel. It does, but this is almost never the failure mode for aluminum - fatigue life is. Keep in mind that "mega scale" stiffness, the stiffness that resists frame flex, and a material's propensity to transmit rather than attenuate high frequency road vibration have little to do with each other. Friends don't let friends ride aluminum.

So if steel is too dense and strong. and aluminum isn't dense enough and is too weak, how about we pick something in the middle - like titanium? Titanium is a great choice, and makes beautiful, corrosion resistant tubes. It makes for expensive bikes though, as it is an expensive material, must be welded in an oxygen-free environment, and is very abrasive for machine tools to fabricate. It also doesn't have the fatigue resistance that steel has, but it's a lot better there than aluminum, so the tubes can be made fairly small and thin for a very nice ride.

Almost every bike you buy today will have a carbon fiber fork. It's not because CF is the easiest material to make forks out of, it's because CF has this amazing ability to kill the high-frequency road vibration that numbs your hands, wrists, butt and feet. For the most part this is because in tension CF is a string, and in compression it is a reinforced piece of plastic tubing made up of epoxy and thread. Plastic doesn't transmit vibration very well at all, and very happily so. (I suspect vibration transmission, or lack thereof, is closely related to a material's efficiency in transmitting sound, but at this point that is just conjecture, not researched fact) My particular bike adds still more vibration damping by adding silicone inserts into the fork, seat-stays and seat post. I also bought a CF handlebar stem and when funds permit, I will be riding a CF handlebar, and perhaps, a CF aerobar.

I knew most of this information when I was looking for my bike courtesy of a neighbor who has been a bike mechanic on and off for 30 years. We went to a LBS because they had a killer deal on a LiteSpeed Ti bike, and he just insisted I look at it. Unfortunately, it was not even close to my size. I noticed the CF forks on all the aluminum and steel bikes and started looking at all CF bikes. I had ridden the Trek Madone 5.2 and it was a revelation, so I asked the salesman to show me the cheapest full CF bike they had knowing I would eventually customize the bike and replace most of the parts. He put me on a Specialized Roubaix and I was hooked.

I ended up buying my Roubaix from an online ad, but had to drive a couple of hundred miles to pick it up, as Specialized, like most of the high-end makers, doesn't want to undercut their dealer network so they are not allowed to sell their bikes online. I had narrowed my search down to the Roubaix or a Felt. I got a great deal on my Roubaix because though it was a NIB bike, it was an '06 model and had a 9-speed gear train. With metal prices so high Specialized had put much cheaper components on the '08 Roubaix of the same model, so I saved almost a grand over a comparable '08 model two levels up. I also knew I'd use the savings to customize the bike with the components I wanted, instead of being stuck with choices Specialized made. Finally, I wanted a triple and not a compact gearing system because we have lots of mountains in California. It was a good fit.

My friend has a custom built Pegoretti, he loves it, and it is a very nice bike, climbs extremely well, and has almost zero flex, but I would not buy another steel bike because metal just beats me up too badly. In fact, I used CF spacers between the top bearing on my headset and the bottom of my CF stem to eliminate any metal to metal vibration pathway. The CF stem, a Ritchey, has an aluminum core, which they really use as a mold, but since it is impractical to remove it after the CF is baked they make you a present of it. With the combination of both aluminum and CF the stem is 3X the strength of an aluminum stem and you don't have to worry the stem will fail from fatigue in the future. It's just ridiculously stiff mounted right down on the headset, and I love it. The wheels are now the weak point in climbing out of the saddle.

This is all well and good, but aluminum is horrible for vibration transmission so before the vibration can migrate from the bottom metal bearing race through the top tube into the top race, and then into the stem, I block it with a 3mm CF spacer. My fork has a CF steering tube, so that vibration path is well damped. An all CF stem and CF handlebars would provide still more damping, but I'm on a budget, so I have to take my hands off the bars occasionally and "thaw them out". You do what you can and live with the rest.

Since I am fighting carpal tunnel I also ride with Fi:zik gel tape and use the excellent Torelli Moda Chunky handlebar tape over that. The Fi:zik handlebar tape is junk, but the gel tape is excellent and can be reused time and time again, although with the Torelli tape, that might be a few years unless you crash.

My next surprise was that bikes no longer come with pedals, and while they come with seats, they shouldn't bother, because all the places where you interface with the bike require very tailored solutions. It took me 4 seats to find one that worked well with me, the chamois in my shorts, and the seat post. As for the pedals, I ditched a perfectly good pair of beautiful old Shimano SPDs in favor of the new wider SPDs that are really a LOOK clone. LOOK had it right, Shimano had it wrong, so Shimano caved and copied.

The old Shimano SPDs have a very small area that contacts the shoe, and after a few hours they tend to create "hot spots" under the balls of your feet. It got bad enough that after a 5 hr ride I was limping, so I shelled out the money for new pedals - Ultegra PD 6620s. Problem solved. The important thing is to get a pedal that has steel on top or you will be buying new pedals every year instead of new plastic cleats.

I also replaced the stock 12-25 cassette gearing in the back, and not for any obvious reason. Shimano cassettes are really just a pile of parts, except that on the higher end cassettes, Ultegra and DuraAce, there are aluminum spiders or carriers that steel gear rings are pinned to. This saves weight, but that is not very important. What is important is the aluminum carriers span the width of 3 gears and 3 spacers so that the contact area between the gears and the splines on the wheel hub is 9 times as great. After one very steep climb I was shocked to find that my large back cassette gears almost tore through their splines and started spinning, ruining the wheel. I replaced that cheap POS with an Ultegra 12-27 cassette and it has performed beautifully ever since.


I have replaced tires, front chainrings, brakes, handlebar tape twice, the chain, the seatbag twice, lights, reflectors and cut 15mm off my steering tube to get my head down in the wind. The Ritchey stem is a 120mm, up from the 100mm stock stem, and I pitched the 20mm tapered headset collar/spacer. All in all I have dropped the handlebar position about 30mm. What I have not done is run out of frame. I still love the frame. It does everything well, and nothing superbly. That makes it very versatile.

My Roubaix still has its plush bike DNA, but it has come a long way in the Tarmac direction since I rolled it out of the bike shop smiling ear to ear. With 31.8mm handlebars low and forward, a stem 3X as strong as normal, and the stem a mere 3mm from the top headset bearing, I ride one of the stiffest front ends on the road - and yet am well protected against vibration. Everything on the bike is aerodynamically clean and while I am not a slave to weight, I pay attention to what I carry - especially in my seat bag. Go small. (The bigger the bag the more junk you'll stick in it) I ride for efficiency and really sweat the details to use my limited physical resources to optimal results.

I rock the hills with the best of them, but the long chainstays and compact frame don't make it as easy as a Cervelo R3-SL would. My triple crank is now sporting 46/38/24 gearing and I can ride straight up 8-25% grade all afternoon in the saddle. I run out of gears over 30mph but the wind drag will eat you alive above those speeds, so it's a trade-off I am happy to live with. I still hit 55 in a tuck going down a half-dozen grades around Sacramento, and love the sure and brutal stopping power of the new SRAM Force brakes. The point is, a bike is a trade-off that is very personal, and most of those trade-offs can be changed so long as you have a good foundation - the frame.