New location

Come on over to my new site:

Going to be posting regularly there.

Thursday, June 28, 2012

Dura Ace Di2 with SRAM XX mountain crankset

Di2 and a custom Seven -- doesn't get much better

Alternate gearing is really popular especially here out west where we have lots of hills to contend with.  The triple crank used to be standard fare, but the advent of the compact crankset, and the improvements it brought (smoother shifting, better chainline, less weight) started the extinction of the triple.

The one downside of a compact, however, is that you can't get gearing quite as low as a triple so some triple-to-compact converts were left wanting more.

The use of mountain bike cassettes with road drivetrains solved this issue, though and the idea isn't new -- when everything was nine speed it was pretty easy to do with the right choice in components.  The switch to 10-speed temporarily inhibited these setups, but work-arounds were found and soon enough SRAM developed their 10-speed mountain group and made it fully compatible with their road group -- a great idea in my mind.  So I had a number of clients do SRAM Red drivetrains (compact cranks, shifters, etc) and use a SRAM XX 11-36 cassette with a SRAM XX derailleur.  Now with their WiFli setup being ubiquitous, this is even easier to do now with many more component choices.

One drawback to using a mountain cassette and a road compact crank, at least from the standpoint of a road rider, is the large steps between gear shifts on an 11-36 or 11-32 cassette; compared to an 11-23/25/28.  A few years ago, I built a custom Seven (Ti/carbon) for a client who wanted the low gearing without the big steps between gears, and we settled on a hybrid solution -- we went with a SRAM Red 11-28 cassette but instead of the SRAM Red compact crankset, we settled on a SRAM XX mountain cranks with the 28-42 chainring combination.  We still included a mountain component but we did it from the opposite side that you'll normally see.

Climbing shifter added

regular Di2 der with 12-27 cassette

This provided him with the low gearing he wanted (and he had little use for the high end gears of a road crankset anyway) and the small steps between gears.  It worked flawlessly and for 3 years he's been putting many thousands of miles on this setup without any difficulty.

Recently he came in to see me with questions about the functionality of Dura Ace's Di2 electronic setup.  I told him the reviews were fairly glowing -- not only does it sport a high "cool" factor, it works exceptionally well.

The question would be could we get his desired gearing on it?

I did some research and found that a few people had gotten it to function with a 32-tooth cassette in back, but this presented the "big gear step" problem as before.  Then I remembered that K-Edge had modified a setup for mountain bikes, calling it Ki2 -- they adapted the rear derailleur by custom designing a new, longer pulley arm.  After looking at it in depth and a phone call or two it was clear they hadn't modified the front derailleur at all, and it was shifting an XTR 2X crankset (which I believe was 26-39).  I already knew they Di2 front derailleur could shift a 16-tooth difference between chainrings (which is standard compact design 34-50) but with my client's 28-42 SRAM XX crankset, we were only dealing with a 14-tooth drop.

Dura Ace 7970 front chainrings are thicker and stiffer than most -- would the SRAM XXs be stiff enough?

SRAM XX was designed to shift ultra-crisp so the rings are 2mm thicker than most and the special bolt attachment (the BCD) is set wider than regular mountain cranks to further stiffen the drivetrain.

I was pretty confident that it would shift perfectly.  But you never can tell with mismatched components, and if we went ahead on this, I was on  the hook for a few thousand dollars worth of components if it didn't work -- not to mention all the work of repeated tearing down and rebuilding with his SRAM Red components if this experiment failed.

Well, long story sorta shorter, is that it works great.  The Di2 front derailleur shifts perfectly over the SRAM XX crankset.  No modifications or calisthenics necessary to get it working.  Check it out:

Even though the bike is all done and waiting for it's owner to return for it, I can't help but go over to it and take the shifting through it paces just to see it work.  About every 10 minutes I've been compulsively returning to it, just because its fun to see a custom project come together so well.  Nerd, I know.

Tuesday, June 26, 2012

Brush Strokes -- Designing and Fitting a Custom Bike

Building a bike, custom or not, is a lot like painting a picture.  You start with very broad, coarse strokes that don't really resemble the end product very much at all except in the broadest sense, and slowly those brush strokes get smaller, more careful and exact and, if you have a skilled painter, eventually you get a beautiful piece of art work.  Anyone can go through the steps, but not everyone has the skills gained from years of training and practice to do them well.

When it comes to painting, I suck.

I can go through the steps -- I can make broad brush strokes to begin with, and refine those to render a painting, but it looks terrible, because it's not something I've trained for or practiced.

In much the same way, anyone can go through the steps of bike fitting or custom bike design, but if they don't have the skills and/or the practice the outcome can be equally ugly.

Building a custom bike is even more tricky because, when done right it is a painting completed by two or more people.  The custom bike fabricators (machinists, welders, finishers, painters, etc) often don't see the rider first-hand.  The rider often just visits their local bike shop who is then supposed to communicate the fitting information to the fabricator.  The problem, and one I've seen a number of times, is that the local shop doesn't really know how to best fit their client, and the task of designing their bike is lost on them, so they leave much of the heavy lifting to the fabricator -- who has never seen the client.  This is when you get a custom bike that at best doesn't take full advantage of the possibilities of the custom build, and at worst (and more common than you think) actually doesn't fit any better than a stock bike off the shelf. 

I recently had a custom bike in the Studio, built elsewhere, in for a fitting and the very expensive titanium frame had the exact same dimensions as the clients previous Cannondale, because those were the dimensions that the shop sent to the custom fabricator.  Now there may be a situation where this would be the best geometry for a client, but that wasn't the case for this client.....that's why they were in to see me for a fitting.

The fitter needs to have the first and last word on the geometry of the bike as it pertains to the fit of the contact points for the rider.  This only makes sense since they are the ones actually seeing the client in the flesh.

In this same vein though, I will defer to the judgment of the fabricator when it comes to angle choices and axle (think about where the wheels end up) placement of the bike.  Good builders have an intimate understanding of the bike they're building -- how the material, tube selection, mitering, butting, welding etc will affect the ride -- and I think they're in the best position to decide if we need to, for instance, slacken a seat angle a quarter of a degree, lengthen/shorten a chainstay a few millimeters, or increase the rake of a fork to achieve a particular trail measurement.  I'm very happy to work with Seven Cycles on the majority of my custom bikes because they excel at this and still will listen to and defer to my opinion on some of these decisions if I can make a compelling case on behalf of my client.

So, on to the process:

1.  The first step is the most 'coarse', where we're determining contact points -- trying to find the most neutral fit position for the saddle and handlebars so that the location of each exists in the middle of an imaginary box.  The box represents the likely and possible alternate positions -- some higher, some lower, some further back, some further forward.  If we can design the frame and then the rest of the bike around these points, so that any point within the box is possible, then we'll have a very easy-to-fit bicycle.

2.  I can take these initial contact points and work backwards to find some of the basic frame dimensions -- I usually start with either effective top tube length and head tube length, or frame stack and reach.  I have some custom excel files I built years ago where I can plug in any combination of head angle, head tube length, stem length and rise, wheel diameter, headset type, headset spacers, seatpost heights, seatpost setbacks, seat angles, etc etc etc and it will give me the dimensions I need.

3.  This bare-bones information is sent to the builder for the first drawing and specs.  From these rudimentary measurements they can flesh out the finer details of the frameset -- the head and seat angles, amount of top tube slope, lengths of all tubes, and not to be forgotten, fork axle to crown, rake.  Another huge benefit of working with Seven is that they build their own custom raked fork in 3 millimeter increments.  Being able to control the position of the front axle is critical to translate the rider's custom positioning into the best possible weight distribution across the length of the bike.  Standard stock forks come in usually just 2 or 3 different rakes, which can severely hamper the handling of a bike.  Ask any of your very tall or very short friends who ride and you'll hear numerous stories about speed wobbles and poorly handling (especially downhill) because the forks for most stock bikes are designed for the middle sizes (53 - 57 cm), not the big and the small.

4.  Once we get the specs back from the builder I can review them with the client and explain what I'd like to change (if anything) and why.  Or I explain why I like the specs, and to the best that I can, I try to "show" the client how these specs will best suit their fit, riding style, and goals for this bike.  I may go back and forth with the builder a couple times until we end up with the specs we're looking for.

As I mentioned before, vetting the specs is critical, and should be the job of the fitter, not the builder since the fitter (me) is going to be the one working one-on-one with the client and the one ultimately responsible for their satisfaction.

5.  Next,  we sign off on specs and the builder begins fabricating the bike

6.  Get the frame, start building.  This is where choosing the correct components is key -- the right crank length, appropriate setback on the seatpost, making sure the length of the stem and the reach of the bars don't cause any problems with the overall reach of the bike that we're looking for. 

This is where I build the bike so that the bars and the saddle get set in the center of that imaginary box -- by properly designing the frame I should be able to get the position here with all the parts being set up "in the middle".  What I mean by this is the saddle is clamped in the middle of the rails by the seatpost, the stem is not overly long or short (maybe 105-110 mm) nor does it have a drastic rise or decline to it (+/- 10 degrees), there aren't a ridiculous amount of spacers under the stem (most carbon forks are limited to a maximum of 35-40 mm of spacers), etc.

If I can do this, then when we go on to the next step, the first fitting, we'll have lots of adjustability in all the settings.

7.  First fitting.  We'll be 90-95% done or more after this fitting.  This goes very smooth since we've done a lot of the heavy lifting before hand and eliminated lots of potential problems because of the way we designed the bike.  The adjustments we have to make are often very small -- a millimeter or two here and there -- precisely because we did plan meticulously ahead of time.

Of course, I still use the capabilities of the infrared system by Retul (every bike I build, custom or not, includes a full bike fit).

8.  Then the client needs to ride about a dozen times.  Often we need time on the bike to fully assess how certain changes will affect the rider.  Muscles and other soft tissues need time to adapt to the new lengths that they're working at. 

9.  Second fitting -- we'll be 98-99% after this fitting.  Because we need time riding to see how the body will adapt, I include follow ups for the first year in the cost of a new bike (and all my fittings for that matter) -- I don't want there to be a barrier, like having to fork out another $50 or more, for my clients checking in to make sure their fit is progressing as we expected.

When clients can't come in for a follow up -- I build quite a few custom bikes for people who drive or fly in from all over -- we're often far enough along into the process that the changes we'll make are known entities (meaning we're often very sure which direction, if any, we're going to have to change things), usually very small, and don't require equipment replacements, only modifications.

A raised seat by a few millimeters; a small tilt of a handlebar; a mild tweak of a cleat.  In the case of clients from out of state, we can often consult remotely (Skype for instance) and work through the change without any trouble.

So there's a lot that goes on in this process.  Sometimes the whole thing can take just a few weeks, other times it can take a few months.  And whether we're building a stock bike or a custom one, the end game is the same -- making sure we get the best fitting, best riding bike for the client.  And the only way we can do that is painstakingly making our brush strokes, one step at a time, until we end up with our masterpiece.


Tuesday, June 12, 2012

Rules of Bike Fitting

This is by no means a comprehensive list, and I'm sure that it'll keep growing and morphing but here you go:

1.  You must use Reason  --  Reason is King.  If you don't have a solid basis for why you're doing something, then you shouldn't be doing it.  I do a lot of fittings for people that have already had a bike fit and are still having difficulties on the bike.  So many of these failed fittings are the result of the fitter making changes to their fit without any real reasoning behind them.  Often they fit a certain way because "that's how we've always done it."  That's not good enough.

2.  Always consider the "whole" while looking at the parts -- Regional Interdependence is the idea that parts of the body (even distant ones) are inextricably linked, and affect one another.  This is a very common theme drilled into every PT student, and one that is especially important when on the bike.  The bike is one of the few tasks where we move with and against a device that we attach to at 5 different points (you can make the argument for 6 contact points if you consider each sit bone, and therefore each side of the pelvis to be separate entities) and when we make a change at one of these points it has the ability to create changes everywhere else.  This is why good bike fitting is so difficult -- a change in foot position is very likely to cause changes even at the rider's shoulder and neck -- reading and trying to predict these outcomes comes only from a deep knowledge base.

3.  A bike fitter needs lots of academic training AND lots of practice -- To successfully combine #1 (Reason) and #2 (Regional Interdependence) requires a lot of education and training, and it's important to realize that these are two different things.  Education is learning all the basics about the bike and the human body (and the human body information is significantly greater in scope) while training is the act of applying this knowledge.

The best scenario would be to have  a lot of both.  There's no accounting for the huge benefit of having thousands of hours of "deep practice" applying what's been learned at multiple years at University.  I remember when I was getting ready to graduate from PT school, my clinical instructor told me that while I knew a lot then (I have forgotten more information than I remember; you get so crammed full of data in PT school) it would take about 4 years of being out there and working to really "figure it out".  I wasn't sure exactly what that meant at the time, but sure enough after about 4 years I started to feel like I could fully trust my instincts when it came to mechanical assessment.  After that, reading someone's movement pattern was something I could do quickly by watching them repeat it (walking, lifting, throwing, whatever) only couple times instead of pouring over video for 10 or 15 minutes.

Interestingly, I discovered, years later, that the four years of practice actually lined up with what researchers have found when studying what it takes to become "expert" at something.  Whatever the task, (shooting a basketball, programming a computer, playing the piano) spending 10,000 hours in dedicated practice of this task can make you as close to an expert as you're likely ever going to be.  Working 4 years at roughly 2500 hours per year -- the math is about right.

4.  Bike Fitting at its core is about Balance --  This can refer to weight distribution on the bike fore-aft -- having the right amount of weight on your hands, feet, and butt -- as well as lateral balance -- most clients are surprised to find out that they don't sit in the center of their bike.

I would argue that this balance can refer to the symmetry and smoothness in a balanced pedal stroke -- one without a hitch or hesitation anywhere through the cycle.

In another vein, the bike fitter must also balance the cyclists physical profile (their flexibility, their strength/stability, their movement patterns) with their positioning on the bike.  One common and very simple example of this that I see is when I have to find the correct bar height that will match the cyclist's pelvic position on the saddle -- once we get the right saddle height and have it's fore-aft adjustment set (we've determined that optimal position of the hips relative to the feet) everyone's pelvic position (and spinal mobility) will dictate where the most balanced place will be for the handlebar placement.

5.  Small corrections!  Bike Fitting is a task of the millimeter, not the inch. -- I am continually amazed at how pronounced even very small changes to contact points can make such a measurable and drastic difference.  Sometimes you need to go big, and when you do, you need to make sure it's not causing other problems, so make sure it is easily undone or walked back if problems arise.

6.  Any bike fit worth it's money needs to be dynamic -- forget the plumb lines, goniometers, and tape measures.  You need to be looked at and have the measuring take place while you're pedaling.  Stopping your pedal stroke to measure -- bad!  Video assessment is fair....using it to "look"  that is, not measure.  The margin of error on measuring with video is on the order of +/-10%, so it's sensitivity is less than exceptional, but a skilled eye may be able to use it to pick out a few problems (see #7 for more).  The infrared used in the Retul system is ideal, since it's dynamic (it measures you while you pedal) and it's hyper-accurate (+/-0.2mm), and measures in all three cardinal planes (thus it renders lateral movements as well) it's the gold standard for measuring.  Notice that I said it was the best way to measure -- having this technology doesn't guarantee a good bike fit.  Since all the Retul does is provide data (very good data), it's still necessary to have a skilled fitter interpret the data.

7.  The trained human eye is a very good but limited tool. -- I'll go on record as saying that it is very important that a good bike fitter does use their own eyeball to help make decisions.  No one should rely entirely on a machine (like the Retul) when performing a bike fit, and the trained eye is great at picking up when something is wrong -- or right.  The trained eye may not be able to tell you what is wrong/right, by how much, and even in what direction something is deviating.  It's great at recognizing patterns or subtle hitches, and at this point in my career one of the times I always use my eye as the final arbiter of when we're done or when we have more work to do.  The machine should never make decisions -- it should only measure a provide data with which to make a decision.

8.  There are no "Have To"s -- There is a lot of bike fitting mis-information out there.  Some written, some just passed along in the verbal tradition from cyclist to cyclist and much of it is pretty bad.

You can see your front hub above/behind your handlebar when you're on the hoods?  Terrible.

You're knee doesn't fall directly over your pedal spindle in the 3 o'clock position?  Ugh, you're in trouble.

These are just two of the many old wives tales and urban legends of the bike fitting world.  There are no absolutes in bike fitting -- sometimes it's in the best interest of the cyclist to have a part of their fit fall outside of the "normal" parameters.  Pick whichever cliche you'd like:  we're all different; no one is completely normal; we're all individuals.  Not everyone falls within three standard deviations of normal, so why do some bike fitters try to wedge everyone into the same box on the bike.

9.  Formulas don't work. -- For the same reasons of individuality as mentioned above, no formula can take into account the vast differences  that exist among cyclists.  Even among similarly sized and built riders, you'll find significantly different optimal fit positions.