First Frame

by Robert Kieffer ()

For many people "building a bike" means going to WalMart or the local department store and buying a box of parts. These are then carted home and, hopefully, after a couple hours of fiddling around with the instructions and bolting the various bits and pieces together, they end up with something reasonably rideable. With a little luck there won't even be too many leftover parts lying around.

However, there exists a subculture to which "building a bike" means something very different. These are artisans who treat the process of building a bike as an artform. They are "framebuilders", and to them building a bike is a very personal experience. Instead of riders having to choose from whatever their local shop has available, a framebuilder will tailor fit the bicycle to each person's specific physique, riding style, and personal tastes, like a finely-made suit. Framebuilder bikes don't start as a box of parts from a store. They begin as a drawing on paper and a small collection of tempered steel tubes which are then cut, filed, shaped, brazed, and welded into something that has tolerances measured in 1/1000ths of an inch.

I recently had the privilege of taking a framebuilder's approach to building my own bike. I designed the geometry, built the frame, choose and assembled all the components, and even created my own paint scheme. It was an intensely rewarding process, but also expensive and time-consuming. Which begs the question, "Why?"

The class I took started with exactly that question:"Why are you here?" Even as I answered, I found my answer both vague and inadequate. What follows starts with my attempt at providing a better one, and then goes on to describe various aspects of the experience that I found interesting. Building a bike from scratch is definitely an daunting experience, so if this is something that interests you I hope you'll find this useful. You are certainly invited to read every word, but I suspect some people may simply find the pictures and sidebar information sufficient.


Starting Out

In pondering why I decided to build my own bike, I realize that I owe much of it to an anonymous criminal in San Francisco. Our garage had been inadvertently left open one day and the Bridgestone XO-3 that I'd been riding for 10 years disappeared. It was a beautiful, fun bike that I only belatedly realized I had completely taken for granted. I'd ridden it for years without thinking too much about it, something I now know is one of the hallmarks of a really good bike. You never think about it, never fixate on the one or two little things that just don't work right no matter what you do. It was only after it was gone, over the course of a frustrating two-year search for a replacement, that I finally appreciated what a unique ride I'd had. It was the frustration of this seach (capped by a broken collar-bone from test riding a poorly designed, and even poorer built production bike) that put me on the rode to enlightenment, or at least awareness.

As time passed, I found myself scrutinizing each bike that passed on the street. Initially this was simply to see if it was my stolen bike, but eventually became a habit that has persisted long after the pain of the theft has gone. Where before I was looking for my beloved bike's color and frame style, I now look to see if what kind of fork a bike has, who the builder is, what the owner has done with the drive train, suspension, etc..

That is how I really came to observe and notice how bicycles are built. To see the little differences in materials, welding and brazing styles, components, and geometries that are all ignored by your average consumer. And this naturally led to critiquing my own rides and riding style, making me a more introspective and aware bike rider.

Some of my observations...

The first and most obvious one is that I'm a freak of nature. At over 6'5", I'm in the top 0.1% of the population. Which means bikes simply are not designed for me. A bicycle that feels just right for an average size person is flimsy and unstable for me. Even with the largest production frames, I have to resort to the usual tomfoolery required to make an ill-fitting frame rideable (longer seatposts, extended handlebar stems, etc.)

I also find that many of the components and materials used today are nowhere near the quality that I have come to expect. Even "high end" components, while technically impressive and high quality, seem over-designed and awkward compared to what was available ten or twenty (or fifty!) years ago. Materials have changed too, and and I find the newer aluminum or carbon fiber frames leave something to be desired. They are soft and squishy, and have a tinniness to them that is disconcerting. Moreover, both these materials have a tendency to fail catastrophically which can lead to rather ugly accidents. Because I tend to break frames, I want something that will fail slowly and that is forgiving by nature.

As I searched for the "perfect bike" I slowly came to the conclusion that what I wanted just wasn't being made anymore. And in fact, probably never had been made. I was going to need something custom built.

But why bother building one myself? The answer to that question is very specific to who I am (or who any framebuilder is, I suppose). I am, by training, nature, and inclination, an engineer. I love to build things and I like to think I have a refined sense of what does and doesn't qualify as fine craftsmanship. This is one reason I love bicycles. A finely built bicycle is inarguably one of the world's most perfect machines. It is the pinnacle of craftsmanship and artistry in a field that has been around for hundreds of years.

There is, however, a small amount of irony in my luddite ramblings; I've made my living as a professional software developer - a field as physically and philosophically different from frame building as it is possible to find. While a bicycle is unique and perfect all by itself, software can only be considered in the context in which it runs. The programs I've written are built upon an ever shifting base of technology. As this base changes, so does the quality of the work. What once may have been a great word processor or spreadsheet is quickly made obsolete and archaic, and eventually non-functional by the constant changes and improvements in the computers and operating systems for which it was built.

As a modern programmer, I define quality as the ability to adapt and be easily maintained (possibly even at the expense of functionality, features and aesthetics). But as a traditional engineer, quality encompasses a notion of timeless form and function. I cannot honestly say if there is a way of synthesizing these two different value systems. I do know that for me the two worlds have always seemed separate.

Regardless, after having spent 16 years in the software world, when I started considering a custom bicycle I knew that it was time to give into the traditionalist in me and take on the task myself.

Finding a Class

To every undertaking there is, of course, a starting point. For me this involved finding someone who could "show me the way". Building a bicycle is no easy feat. Even ignoring the substantial skill and expertise needed, it also requires tools, materials, and techniques that are not readily available to a novice. Thus, I needed a teacher.

Finding one was not all that difficult. But I first needed to disabuse myself of any romantic notions about apprenticing under some master frame builder, or even a not-so-master. A few calls to some builders I had contacted (or at least knew of) sufficed to verify that, although nice enough, theytended to be too busy actually building bikes to deal with someone who didn't know diddly about their business.

So, I found myself looking for schools that offered classes. The list of candidates turned out to be pretty short - there simply aren't that many schools that offer bike building courses. Actually, there was only one. The United Bicycle Institute (UBI) offered a class in traditional lugged frame design and construction - the technique I was interested in.

A few phone calls later and I was signed up.

The Build

So it was that in May of 2004 I found myself sitting next to 7 other aspiring frame builders faced with the, "Why are you here?" question. As I looked around and listened to the various answers, I really enjoyed the sense of displacement I felt. No longer was I surrounded by people fixated on the hi-tech world of software, but rather I was with people who shared the same traditional values that I'd been suppressing for so long. My fellow students came from as far away as New York and Louisiana, and from backgrounds that were, well, almost alien in their strange unexpectedness. I could only admire these folks who had walked a path so different from the mundane and comfortable route my own life had taken.

There was a certain theme among us however. Of the eight of us, two were bike mechanics, one a bike messenger, and two more were bike shop owners. That left just me and one other software guy from Silicon Valley. I suppose us two geeks had lives that looked equally strange to them. We do, after all, spend the majority of our day just sitting at a desk. (*sigh*)

Yet, to a tee they were all fantastic individuals. I couldn't imagine anyone else I would rather have spent those two weeks with. Not only were we shoulder-to-shoulder in the classroom 40 hours a week, but for the six of us living in the Cycle Hostel we were pretty much living on top of each other the rest of the time as well. It made for a memorable experience and I'm certain I'll remember our motley crew for a long time to come.


I found the class both fascinating and frustrating on many levels. Surprisingly, the actual design of the frame was completed in under a day. UBI provides a guide that simplifies this process immensely. Furthermore, part of the prep work you do before the class has you figuring out some of the basic dimensions that are likely to work well for you based on various body measurements and measurements of bikes you've ridden. So we only spent the afternoon of the first day and the morning of the second coming up with our full scale drawings.


The transition from design to construction was a little nerve-wracking for me. My goal was to build a lugged mountain bike frame, but lugs are notorious for only working with certain geometries and the frame I wanted was not normal by any means. Going into the class, I knew there was a pretty good chance that once I'd laid out the design I would discover that there simply weren't any suitable lugs available. It turned out to be a very close call. The lugs I finally used would have to pushed almost three times as far as the "comfortable" tolerance they were designed for. But the important thing was that I had a lug set, and from that a frame would follow... in theory.

With the design out of the way, it was time to address one last minor issue: none of us had actually done any real torch work before coming to this class. Okay, that wasn't strictly true - several of us had done some work, including one woman with quite extensive experience - but by and large most of us had little experience brazing anything, let alone rather temperamental bicycle hardware. Thus, we spent the better part of a day practicing filing head and seat lugs, then brazing them onto whatever pieces of scrap tubing happened to be lying around.

Eventually Ron, our teacher, had us get down to the business of actually attaching the head tube to our top tube, the first actual construction required for our frames. I'm sure all of us would have been happy to spend a little more time burning our practice joints to a crisp, but at day two time was already of the essence and we needed to take the leap of faith.

And so it was that we started the process of filing, sanding, filing, brazing, filing, cutting, filing, filing, filing...

Mitering and filing

Did I mention we did a lot of filing? It turns out this is one of the quintessential skills of a framebuilder: Creating a strong joint starts with making sure the components fit together properly. For a lugged joint, this means not only making sure that one tube fits snuggly against another tube, but also making sure that the lug which holds the two together is sufficiently loose. The latter may seem counterintuitive, but a tight lug joint creates all sorts of problems. To work properly, the filler material (in our case, silver) must flow completely into the joint and create a strong bond, which is only possible if there's enough room for it to flow. It's not uncommon to hear framebuilders debate whether or not a particular kind of filler needs a .06" or a .05" gap. And developing the ability to create this kind of fit by hand is just the first of many skills a framebuilder has to learn.

By the end of the class, I sort of had it but it came at the cost of about 20 hours of filing and more than one bruised and battered finger.


This was actually the most thrilling part of the whole process. The flame, the stench, the tension! Once you have a joint properly fitted (or so you hope) and everything all set up in the frame jig, its time to put it all on the line and fire that sucker up. You start by applying a healthy dose of flux, a mildly acidic paste that cleans a joint, and seals it off from air once heated. Then you start applying heat. At some magic moment, when everything is nice and hot, you add a dab of silver and watch as it disappears into the joint, or crawls along a seam, or decides to go in some random unexpected direction. It is this moment that metalworkers talk about when they say metal has a mind of it's own. Heat does weird things as joints expand, and molten silver or brass seems to take on a life of it's own.

This is also where the framebuilder's efforts are won or lost. Once a joint is brought up to temperature, the working time available is measured in minutes, and either you join it properly or you're left with a charred mess that has all the aesthetic and mechanical properties of charcoal briquette. Knowing how and when to apply heat and filler is vital.


One of the side-projects of the class was fabricating a fork for our frames. I already had a suspension fork for my bike so I didn't really need a handmade fork. However, I really wanted to try my hand at this and I figured it'd be fun to have a rigid fork for the bike if I ever decided to use it. What made it especially fun was that I got to work through some of the frame geometry issues involved in adapting a rigid fork to a frame designed for a suspension front end. As it turned out, by making a fork designed for a 29" wheel, I kept the angle and handling characteristics of the bike reasonably unchanged.

There's not a whole lot to say about this other than it's just very satisfying to hand build a fork to go with your frame. Even though I have a really nice suspension fork which works just great, in my mind's eye I still imagine my bike with the lovely lugged crown and dropouts on the fork I built.

The Problem Lug

The biggest stumbling block of my build occurred when trying to fit the down tube to the head tube. As I mentioned earlier, I was using a lug set that wasn't particularly well-suited to the geometry of my frame. The result was that it took a lot of effort to get most of the lugs to fit properly and this joint in particular was giving me a lot of problems. I spent the better part of an entire morning filling and fitting trying to achieve the 74° angle I needed. Each time I would fit the tubes together to see how close I was, I would discover that the joint was still tight and yet more material needed to be removed. Finally, I felt I had it close enough that all I needed to do was flex the joint just a tad to loosen it up and I'd be golden. As I gave it the gentlest of nudges, I heard the unmistakable *snap* of the lug giving way. Sure enough, I'd cracked the lug right across the top. My heart sank, not just at the loss of a morning's worth of work, but also at the prospect that there was only one other lug and I was not at all sure I would be any more successful with it.

My salvation came in the form of Keith Anderson, a framebuilder of some renown who had just joined the faculty of UBI. At his suggestion instead of throwing the cracked lug away, I actually used the crack to give me the amount of flex I needed. Sure, there was a crack there, he said, but now you have a lug with the correct angle!

"But it's still broken!", says I.

"No problem, just fillet braze the lug."

*Wow* now why didn't I think of that? Well, because I didn't have 25 years experience among other things. Needless to say I was absolutely elated with this solution. Not only did it salvage the lug and my morning's work, but it gave me the opportunity to try my hand at a fillet joint, something I wouldn't have otherwise done on my lugged frame. After a little brass brazing and some filing to clean up the fillet I had a perfectly useable lug. The final product worked beautifully and as you can see from the picture on the right it's basically indistinguishable from a normal lug. And any concerns I had over whether or not the joint would be strong enough were a) Pshaw'ed by Keith and b) later proven groundless in light of the beating I've given this bike with absolutely no sign of weakness showing.

Anyhow, like the master japanese potter who knocks a hole in each of his works because it is a sin against god to strive for perfection, my bike, too, has a hidden flaw that affirms it's construction by the hands of a mortal. :-)

The Head Badge

The "signature" element of any bicycle is the head badge - the small decoration or design that ornaments the head tube at the very front of the bicycle. A lot of builders use a decal or some painted design here, but the traditional approach is to use a small metal badge that is riveted in place.

Our options in class along these lines were limited to, "whatever we could cobble together in our spare time". There wasn't any time formally set aside for doing this so we students were left to our own devices. Most elected to simply forego the process, but I felt it was important on my bike (especially because I had such a long frickin' head tube!). Unfortunately I hadn't really thought ahead about what I wanted.

I toyed with a number of ideas: a little dog, a bone, a moon, something. All I knew was that it had to be simple because I didn't have the tools or time to produce an elaborate design. I spent the first week of class pondering what I wanted to do when finally Christi, one of my fellow students, came to my rescue. She'd mentioned that whatever I, it should probably be made of stainless steel, and followed up on this a few days later by supplying me with the an old sink disposal ring that she'd found in the trash. It conveniently had the words "KENMORE DISPOSER" embossed on it which were surely good for something, right?

It wasn't until about mid-way through our second week that inspiration finally struck. I took my mini hacksaw and file (I was doing this on the couch in the hostel) and chopped out the R, M, and K letters. By flipping the M upside down to form a W, I now had, "RWK" - my initials. The next day, these were quickly brazed on and my first head badge was formed in all it's dubious glory!

The Last Days

My memory of the last couple days of class are somewhat of a blur. It was apparent early on that I was going to be pressed for time, and so I pretty much spent every minute I could in the shop the entire two weeks I was there. By the time we got to the final step in construction, the "braze-ons", my exhaustion and nervous energy had me in a jittery fugue state that went well with the thunderheads that were building outside the classroom. "Braze-on", it turns out is a catch-all term for all the niggly little crap you have to add to a frame to actually make it useable. Things like bridges on the chain and seat stays; braze-ons for racks, water bottles, cable guides, and tire pumps; Front and rear brake bosses. A barebones track bike will need only a handful of extra bits, but I wanted to outfit my bike for touring which meant that I pretty much had to put everything on it.

And so my final hours were spent drilling and brazing like a mad-man. When it was all said and done I had something that looked like a frame, smelled like a frame... but would it work? I honestly didn't know. Would I go to mount the cranks only to discover they bashed into the chainstay? Would the front wheel rub the down tube? Had I faced the head tube and bottom bracket straight? I was genuinely worried that I'd miscalculated something and would only discover it when I was too far away from the comfort and safety of the classroom to be able to fix it properly.

Yet, by the time I'd cleaned up my workbench, and gotten the worst of the grime off my frame, I was elated. I'd done it, I felt like I'd actually built an honest-to-God bicycle.

Of course, the elation wore off pretty quickly and I realized I didn't have anything I could ride yet.


As it turns out, fabricating a frame is only the first of three fairly major steps in building a bike. The second step is getting it painted. This is necessary not just for cosmetic reasons, but also because even the highest quality frame will quickly rust into oblivion unless it has some protective coating.

At the time I finished fabricating my frame, I didn't really know what I wanted to do for the finish. My options were homebrew, wet paint, or powder coat. The homebrew solution amounted to little more than, "take the frame out back and hit it with a couple cans of spray paint". While this appealed to me on an economic level, I'd put way too much time and money into this project to settle for the rather shoddy finish I was likely to end up with if I did it myself.

Instead I looked to a professional finish, either powdercoating or the more traditional paint technique. There are pros and cons to both approaches, but the deciding factor for me was a combination of cost, suitability, and convenience. Prior to going to the UBI class, I'd done some research to help figure out what I wanted. While I'm not a graphic artist, I do know my way around PhotoShop. And so I was able to grab a generic picture of a frame off the web and use that as the basis for my design work. I was pretty sure that I wanted to stick with a 2 or 3 color scheme at most, so my designs didn't get too elaborate. Here are a couple of the mockups I did:

On the advice of a friend who was building a chopper, I contacted a local graphic artist who had a pretty good reputation. I went by his shop and saw several examples of his work - surfboards, motorcycles, cars, even mailboxes - and he was obviously a talented artist. Even better, the price he quoted me on my "expensive" option (the middle one above) seemed reasonable.

And this is where we got into trouble. In retrospect, I think his enthusiasm for trying something new led him to underbid the project. By a lot. My bike featured a 2-color fade, pearl white "flaming" panels and lugs, and custom decals, all of which he said he'd be able to do for $300. It was only later I would discover that he probably should've charged me at least double that.

Great, I got a deal, right? Well, not really. By the time I actually had a frame for him to paint, his enthusiasm had worn off and he'd managed to completely swamp himself with work. This in spite of the fact that I had scheduled the time to do this two months prior. He had also come to the realization that it was a lot more work than he had originally anticipated, and basically told me he just couldn't do it.

This put me in a really tight spot. I was planning to do a ride on this bike that required me to leave about a month later. In calling around to other painters I found that most simply couldn't fit me into their schedules, and the ones that could were asking two or three times the price, plus shipping, plus rush fees. Needless to say, I was pretty upset.

As a last resort, I went back to my painter and told him the bind he'd put me in. Fortunately I managed not to get too irate, and we worked out a compromise where he'd try to squeeze me in and I'd pay a little more. But even this didn't work out as well as I would have liked. He was so pressed for time, and margin of profit so slim on my bike, that the quality suffered. Nothing major, but there were a number of little "issues". He failed to apply paint to a couple of crucial areas (outside face of bottom bracket, top of seat tube), one of the panels was oriented the wrong direction, and a decal wasn't applied properly.

Needless to say it was more than a little frustrating. With hindsight, I realize that my expectations were probably more than a little high. I had a significant emotional investment at this point and wanted everything to be ... just ... so. Expecting someone else to have that same passion and attention to detail was probably a little unrealistic. I don't feel like I got ripped off - As you can see, the frame looks really nice - but come the day I decide to turn the bike into a 29" fixed gear, I won't spill any tears over stripping it off and doing something different.

Putting it Together

Holy cow, this writeup is turning into a tome! Much like the actual build, I didn't expect there to be this much to cover, so I'm gonna try and wrap this up. The step to getting a complete bicycle is the final assembly. You know, putting on the brakes, wheels, handlebars, etc...

In some ways, I found this to be the most challenging aspects of this whole process. If you remember my little rant at the beginning, I'm not a big fan of modern components. For this bike, I had some very specific ideas about how I wanted it outfitted. I needed fairly rugged components to handle the touring I was planning on doing, and I was set on using a 7-speed drive train that hadn't made in 10 years. Without going into detail, I spent several weeks tracking down all the components I was after. In the end, I found myself mixing and matching several generations of parts, which I got from at least 5 different sources (including one in Korea!)

But, the day finally came when I had everything I needed. The actual build of the bike went rapidly after that. Much to my surprise (and sincere relief), I found that I had actually managed to outfit the frame with all the brazeons and other doodads I needed and put them where they needed to be. There was no need to go back to the workbench and file or sand or light up the torch again.

The First Ride(s)

My first ride on the new bike turned out to be rather anticlimactic. When I had first conceived of this little adventure, I imagined myself donning my cycling gear and heading out into the pristine wilderness for a long afternoon ride. You know, the gentle vibration of the road massaging in my arms, the smooth hum of the chain, the wind caressing what's left of my hair, blah blah blah.

As it turned out, I finished the bike about 10 minutes before my wife and I were supposed to go to dinner with some friends, which put me in somewhat of a quandry. While I wanted the first ride to be special - I had, after all, put a lot of time into this moment - I also really needed to know whether or not I'd built something worth riding one mile, let alone hundreds or thousands.

In the end my anxiety won out and I decided to go for a quick spin around the block for a sanity check. The first couple of minutes were spent getting a feel for the cockpit ("a tad narrow"), the brakes ("check"), the shifting ("pretty good"), and the new suspension fork ("sweet!"). And after that it was just a matter of seeing how the bike felt as a whole. I tried swerving back and forth a few times to see how it recovered, went no-hands a bit to check the stability and alignment, mashed on the brakes to test stopping power... and... well... it seemed okay. The bike rode well, was plenty stable, and had a really nice feel to it. So actually it was more than okay... it was pretty darn good... really good, come to think about it....

... and that was how quickly it happened. In a matter of moments, all the anxiety and doubts disappeared, all the time and hard work paid off. I found myself on a fabulous bike that I couldn't wait to get out on the trail.


Its been about 6 months since that first test ride. The bike currently sits in the garage waiting for the snow to melt so I can take it out again. I've probably put about 300 miles on it. Which is not a lot by most standards, but virtually all of those miles are on the singletrack in and around Bend, Oregon. These are trails that challenge both bike and rider, and I've done my best to push things to the limit. In the beginning I would wonder whether or not I should be jumping as high or bombing downhill as fast as I was, wondering what might break because I didn't do quite as good a job of something as I should have. But the bike has held up just fine.

At this point the weakest component is me, the rider. I doubt my skills will ever match up to the kind of punishment this bike could take, which is a very refreshing thought. The best part about all of this is that after a hairy section of trail - the teeth-rattling downhill on "Flagline" or the endless string of jumps on "Whoops" - I find myself grinning like a fool as I realize yet again that this is my bike I'm riding.

Robert Kieffer
Jan 19, 2005

Recommended Reading

These are a couple articles I stumbled across that may be of interest: