Building wheels

Past experience

I've been messing around mechanically with bikes for decades. I learned the basics in childhood doing all the basic tuning and repair activities. But it is only in recent years that I've started to take on the full spectrum of work. Recently, I've started building bikes from the frame up (Litespeed Road, Litespeed Gravel, Waterford) and fully overhauling old bikes with various generations of technology. Doing this brought in new skills like replacing headsets, working press-fit bottom brackets and hub bearings, hydraulics, and more.

But wheelbuilding is scary

But there was one major component build that remained scary voodoo to me: Wheelbuilding. I'd been able to true a wheel without screwing it up since I was a kid. But the idea of starting with a hub, a rim, a stack of spokes and a pile of nipples made me very uncomfortable and I avoided it.

I avoided it even when the rear wheel of my commuter bike made it clear that it had been built incorrectly by breaking spokes every few months. I just kept replacing them until I got lucky and 'solved the problem' when I found an 11-speed Alfine wheel to upgrade it with.

Backing in to it

But then, I was forced to rebuild the Ibis wheels on the Litespeed Gravel. They were fundamentally good wheels with decent hubs and carbon fiber rims. Unfortunately, a previous owner had probably allowed ammonia-based sealant to get to the nipples. They were aluminum and all corroded. This led to several sequential spoke breaks. I decided to bite the bullet and realize they needed a rebuild with new nipples and spokes.

But I figured out a 'cheat' to avoid the 'danger zone' of a full build. I replaced each spoke, one at a time. This kept the wheel structured the same throughout the process. No mystery to lacing. No worries about ending up badly dished or out of true. No unknown spoke lengths. This was not a fast process, but it worked well and resulted in wheels that have now run for over 700, often off-road, miles (they are on a gravel bike after all). 

Finally, a need

Now we get to building the Waterford Adventure Cycle. I was having trouble finding wheels that fit the unusual size configuration and were sturdy. They needed to be MTB width QR hubs at 135mm rear spacing but with 700c rims. Most MTB QR hubs for sale are laced to 26" rims. And they need to be beefy to carry a touring load and this clydesdale rider. I tried an old cyclocross wheelset for a while, but got stranded on my very first loaded adventure when a spoke broke (to be fair, the wheel did get caught in a road crack).

Fortunately, the experience with the Ibis wheels gave me the courage to go through with it and build them up from parts.

First, learn some things

Before tackling the project, I wanted to understand both the theory and practice of wheelbuilding. At some levels, wheels are simple. Balance the tension around the rim to keep the wheel true. Check the overall tension to make sure it is 'enough' but not 'too much'. But the details are vastly more complex. 

For example:

• What parts to use? Hubs and rims are the obvious. For hubs, it's really just a matter of picking one with the right price/quality match along with the right number of spoke holes. Rims are a little harder because there are many different designs but at the end of the day, match the spoke count and features and you're there. Then one must select the spokes and nipples. What material? What gauge (thickness)? Butted or not? Aluminum alloy or brass nipples?

• How long do the spokes need to be? When replacing a spoke, one can usually just measure the old one and know it's right. When you need to start from scratch, it is complex geometric math related to the diameter of the circle where the spokes enter the hub, the actual diameter of the rim at the base of the rim where the nipple sits, and the angles produced by both the lacing pattern and the offset of the hub flanges due to the freewheel and sometimes the disk rotor.

• What is the right tension? If spokes are overtensioned, they break. Duh. But non-intuitively, if they are not tensioned enough, they break too, from fatigue. When the wheel is 'too soft', each spoke stretches and moves too much with each rotation and eventually fails from metal fatigue.

• What lacing pattern? Spokes are usually 'crossed' 1, 2, 3 or sometimes more times as they run from the hub to the rim. This gives some additional strength for rotational torque, along with better characteristics as they flex during rotation. Raidal lacing (no cross, straight from the hub to the rim) is done. But it is not recommended for heavy-duty applications. 

So many questions, not enough knowledge. So I went on a research mission.

Resources

The basics

If you want to take this on, I recommend a few resources if you want to be well-educated. 

The first is an excellent, borderline-academic book on the whole concept of wheels: "The Bicycle Wheel" by Jobst Brandt. It is remarkably hard to find since it was from a small publisher, and Jobst Brandt died several years ago. 

Try to find it in a library if you're lucky (I wasn't). Purchased copies tend to be silly expensive. I managed to find a decently priced used one online eventually. Today I see a store that appears to have reprints for only $30. That's cheaper than a lot of used copies. I can't vouch for the seller, but they look legit. 

It is a fairly technically dense book, but it is also a quick read. Start here with your studies for foundational theory, along with proposals for good methods of assembly.

There are various written online sources of information. Some are good. But most are not as thorough. If you really want to learn, I highly recommend the book as step one.

The practical

From there, I recommend video and supplemental online resources. These will let you see what needs to happen. For most of us, that will save a lot of dumb errors when trying in reality. Park Tools really has an excellent set of videos on the topic that walk you through and show the process. Here are a few highlights

• For actually building the wheel, I think this is the best one, by far. Clearly presented, step by step. You can even watch a segment, do a step, and repeat until finished.

Note: Calvin presents a method of building a 3-cross wheel that is different from the build chapter in Jobst's book. They get to the same place. It's just a 'tricks' difference. Decide which one resonates with you the most and use it. FWIW, I went with the video version because I liked being able to have visuals to walk through as I built the wheels.

• For some background on lacing decisions, two to watch:

• A blog post to determine ERD (Effective Rim Diameter) needed to calculate spoke length: 

• To learn about calculating spoke length:

Calculating spoke length 

I learned that it is not as straightforward as one would hope to determine the right spoke lengths. For some collections of parts, you can find a calculator that accurately knows the dimensions of the hubs and rims and can calculate it for you. But they don't have the data on all of them. So you need to go through a process:

1. Determine the hub and rim you wish to use
2. Look up the specifications/diagrams for them. 
3. Get the hub and rim in your hands and measure yourself. If at all possible, do this before you order spokes. Things like the flange offsets are listed differently with different manufacturers and used differently with different calculators. Having the real measurements will both help cross-check and make sure you are using the right measurements.
4. Use more than one calculator. For various reasons, they won't all give you the same answer. The 'right' answer can vary by a millimeter or two without harm. Part of the variance is different opinions about rounding methods. But you don't want to be far off. I was initially (before I could measure my parts) getting significantly different answers.

Calculators

• DT Swiss (simple if you're using their parts): https://spokes-calculator.dtswiss.com/en/
• Sapim (spoke vendor): https://www.sapim.be/spoke-calculator
• Pro Wheelbuilder (has diagrams): https://www.prowheelbuilder.com/spokelengthcalculator
• United Bicycle Institute: https://bikeschool.com/resources/spoke-calculator/
• SpokeCalc: https://spokecalc.io/

NB: See #4 above. Use more than one and make sure you get a compatible answer.

Additional Resources:

• Wheelbuilder group on Facebook
• Online tool (Park Tool) to track the tension on each spoke as you build the wheel and calculate the variance tolerances. This is a good way to make sure you're getting good results.

Tools needed

You do need to own or borrow a few specialized tools to do this well. There are very expensive versions of many that are made for professionals in shops. They are better. But you don't necessarily need to spend that for building one or two wheels. You can make up for some of their features with patience and tricks.

• Standard spoke wrench: Do not skimp here. Buy a quality wrench from a quality vendor in the size you need for your build. They aren't that expensive and cheap ones fit poorly and will strip nipples. Most work from the 'inward side' (toward the hub) of the rim on square flats of the nipples. 
• Nipple driver: This one is optional since you can use a screwdriver and/or the spoke wrench, but it can be super helpful. Conventional nipples have a flat screwdriver head on the nipple that you can access from the 'inside' of the rim (tire side). These drivers are a strange, bent-shaft screwdriver on a swivel that makes it easy to quickly drive the nipple down the spoke until it reaches the baseline point.
• Inside spoke/nipple wrench: (option vs above) Some nipples use square or special (squorx for example) heads inside the rim. These are much easier and more reliable to drive down than the flats on conventional nipples. You can usually use this end to achieve final tension as well. But you may not be getting that kind of nipple for other feature or price reasons. They are great if you can. And get the matching tool.
• Truing stand: While you can fairly easily use the bike frame to true up a wobbling rim, I argue that building a wheel while using the frame/fork for reference is a recipe for mental pain. You can get away with a cheap, simple stand, though. The shop-quality stands are more sturdy and stable. They also make it much easier to measure the runout, etc. But a basic one that locks in the hub properly and can hold a pointer in the right place is adequate. You could probably make one yourself that would be adequate.
• Spoke tension tool: Experienced builders can do a decent job, literally tuning by finger pulls and sound. Neither you nor I are experienced. Get a tension measuring tool. Don't skimp on a cheap one. I had a cheap one for some time, but decided to invest in a certified, calibrated Park Tool one. When I cross-checked them on an existing wheel, I found the cheaper one was off by a lot. Not great. Like the other expensive tools, borrow or rent them if you are able. But here you need to make sure it has been calibrated and not damaged. This is important because you generally want tension to be close to the limit for the weakest link (rim or spoke) without exceeding the limit. Guessing too low and the wheel will be weak. Guess too high and you will overload parts and break them eventually.
• Dishing tool: Recommended but optional. This is a simple tool that is relatively expensive. It allows you to easily determine the distance from the plane of the rim to the hub. It is used to make sure that you have centered the rim relative to the hub. This is particularly tricky with a rear wheel  (or a front with disk brakes) because you build with different length spokes on each side at different tensions. You CAN get away with using the bike or other hacks to do this. But a dishing tool makes it easier.

My Results

My first wheelset was for the Waterford. I made all the parts decisions based on being sturdy rather than light and keeping costs low.

• Rear Hub: Shimano FH-M580 Deore LX 36h. NOS 9-speed, QR MTB hub that matched my 135mm spacing and 3x9 drivetrain. Quality reputation at a low price
• Front Hub: Shimano HB-RS400 36h. A more modern, inexpensive, but reputable hub.
• Rims: Ryde Sputnik 36h. Double wall, double eyelet, sturdy rims widely reported for years as the best touring rims. A little tricky to source in the USA, but available.
• Spokes: DT Competition Double Butted 14 gauge silver. These are well-respected, conventional spokes. Double butted because after reading Jobst, I came to understand that a) Spokes tend to break at the nipple or J-bend where double-butted spokes remain thick. b) The thinner butted section makes for better spoke elasticity as the wheel loads in rotation or impact. This deflects load from the vulnerable ends. It also makes it easier to get the right tension preload. Non-butted spokes have a theoretical breaking strength that is higher, but not where it counts and the thin section is plenty strong anyway.
• Nipples: DT Swiss silvered brass. Yeah, brass nipples weigh more. On this wheelset, it doesn't matter. But what is important is that they don't corrode easily and remain serviceable. Silver to match the bike's look.

Calculating spoke length was the worst part for me. Before holding the parts, different online part specs were coming up with different answers. (Remember, don't trust them without checking). Once I had the parts in hand and measured them, I got much closer results. 3 of 4 calculators rounded to the same answer and the 4th wasn't much different. When I built the wheel, the measurements were spot on.

In the end, I got a nice, sturdy set of wheels at a very reasonable price with the only real downside being the additional weight. The new rims were a couple of mm wider, which gave the tires a better shape. From the first roll, they have held dead true for 612 miles (mostly laden with 40+ lbs of cargo and a clydesdale rider) as of this writing. The original/placeholder wheels were lighter, though. The new ones added 610g for the set, which is non-trivial even for this non-weight-weenie. But then again, the previous set broke a spoke on the first loaded ride.