The Personal Web Pages of
Chris X. Edwards

Chris' Wheelbuilding Guide



Bicycle wheels are ingenious devices, but they are not as awesomely complex as most people believe. Well, they don't have to be. With a little patience and interest, it is possible for anyone to understand the practical elements of the bicycle wheel.

There are many little tricks to keep in mind. The most important one is to look at other wheels. If you have a pile of spokes, a hub, and a rim, the simplest way to turn that into a wheel is to copy a successful wheel. Another thing to consider is that the complexity of bike wheels is cut in half if one considers only one side at a time. By considering only the port or starboard side of the wheel (left/right), you can simplify things quite a bit.

One particularly nasty aspect to building your own wheels from scratch is selecting the correct spoke length. This task is not as formidible as it seems. I've very explicitly and thoroughly explained not only how to do it, but the thinking behind the process. Don't get too discouraged if you get lost in the theory of where the formulas come from. They are still easy to use.


The first step in custom wheel building is collecting information. If you walk into a bike shop and ask for some spokes, the first question is going to be "How long?". That depends on two things - the rim and the hub. There is a lot of variation between different hubs and rims in the world and rather than consulting a general table of this information (if you can even find such a thing), it is safest to make the measurements yourself.

Rim radius is one of the more tricky things to measure. There may be some kind of statistics from various rim manufacturers, but I would be skeptical about exactly what was being measured. I like to do it myself and here's my preferred method. I take a piece of wire and I wrap it around the entire circumference of the rim where the rim strip or tape usually sits. Then I take some pliers and twist the ends of the wire to tighten it taut around the rim and to hold it in place very snugly. With it held in place, you can adjust the wire so that it runs around the rim directly over each spoke hole at the center of the rim. If the rim is double walled, then I slip a spoke nipple into one of the holes. Then I measure from the wire to the top of the nipple. This is the basis for a rim measurement that relates to the useful task of sizing spokes. Then I use some wire snips and cut the wire a few cm from the twisted connection. This wire is now a very accurate representation of the circumference of the rim at a point with a known reference to your spoke. Measure this wire and divide by PI to get the diameter and further divide by 2 to get the Rr.

This is one of the critical measurments of the hub that makes this whole custom spoke length necessary. The "bolt circle diameter" in machining terminology describes this property and can be easily measured with calipers or dividers (or a bent paperclip and a ruler). The distance from one hole to the hole on the opposite side of the same hub flange (180 degrees away) is the BC diameter.

Flange Offset (z1)-
z1 is not so easy to find out, but it is simple in concept. z1 is simply the distance of the hub flange from the center of the wheel. How would you know this? On front wheels, it is easy: simply measure from the center of one flange to the center of another and divide by two. This simplification is based on the fact that, generally, front wheels are symmetrical laterally. Rear wheels pose a bit more of a problem. In fact, often rear wheels have two different sizes of spokes, a port set and a starboard set. The important information is where the flanges of the hub are with respect to the center of the dropouts (the frame). What I usually do to measure this is to set the hub up vertically with the part of the hub that normally touches the inside of the frame resting on two blocks (cassette tape cases work). Then I can measure from each flange to the blocks. If you know the distance to each flange and the dropout spacing (usually 135mm for mountain bike frames), then you can easily figure out the distance from the theoretical center of the hub to each flange (z1). Don't forget that each side may be asymmetrical and you might even need two different spoke lengths for a back wheel.

It is also worth noting that for fussy folks, measuring the flange offset merely to the general center of the flange may not be high-tech enough for you. In that case, you need to do a calculation for each side of each flange. So if your flange is 3mm thick and 30mm from the center, you may need to use a z1 of 32.5 and 27.5 (don't forget that I'm adding the radius of a 2mm spoke!) If you're not a fussy person, you can safely ignore this level of detail.

Spoke Length Calculation

I used to think this was a very complex topic until I needed to calculate some spoke lengths by hand. After thinking about it, I realized that the problem is not very difficult at all and I had imagined complexity that wasn't really there.

Basically, to find the length of a spoke, you must use the same mathmatical technique that you use to find the length of anything: the distance formula. This is based on the simple and well know fact that X^2 + Y^2 = H^2. (For example, a right triangle with side lengths of 3, 4, and 5, shows this with 9 + 16 = 25).

By rearranging this formula, the hypotenuse (H) of a right triangle can be found with:
H = sqrt( X^2 + Y^2 )
With a bit of thought, it is possible to see that a third dimension can be added like this:
H = sqrt( X^2 + Y^2 + Z^2 )
There is a further refinement that can be made to simplify calculations. Each component (X, Y, and Z) is the distance in a particular direction from a start location (the spoke hole in the hub) to an end location (the end of the spoke). This means that the zero point is on one end of the spoke. This isn't very convenient and it can be moved easily enough by subtracting the spoke endpoint coordinates to get the lengths independent of reference points. This turns out looking exactly like a classic distance formula:
Spoke Length = sqrt[ ( x2 - x1 )^2 + ( y2 - y1 )^2 + ( z2 - z1 )^2 ]

So how does this relate to the spoke length? Basically, if you can find out how far your spokes need to stretch in each of three planes (horizontal, vertical and axial), you can find the length for the final spoke. This sounds more like a complication than a simplification. Now we need to find three lengths to find one, right? Yes, but each of these three lengths is much easier to find than the overall length.

All that is necessary to do is to find the locations for all of the variables in the distance formula. First, we need to choose an orientation. Let's call the center of the wheel the 0,0,0 point. Looking at the wheel from the side along the axle, we can consider the positive x axis going from the axle to a hole on the rim out to the right. The positive y axis will go from the axle to a hole above the hub. Remember that the hole of the x axis is separated by the hole of the y axis by 90 degrees or 1/4 of the wheel. For a 36 spoke wheel, move to the next spoke over 9 times. The final axis is the z axis which runs along the axle. With distance calculations, the positive and negative aren't really too important as long as you're consistent, but generally, if the positive x is to your right and positive y is up, then positive z is toward you and negative z is away.

Now that we have a good coordinate system to describe the wheel, let's look at some subtle things that are conceptually rather significant. The first question is how exactly does the rim line up with the hub? One could imagine that based on the symmetry involved, that the spoke holes of the rim will line up in some way with the spoke holes on the hub. There are three possibilities. Possibility one is that if a spoke hole on the hub lies on the x axis, then there will also be a spoke hole on the rim where it meets the x axis. The second and reasonable possibility is that if there is a spoke hole on the hub at the x axis, then the rim will intersect the x axis exactly between spoke holes. The third possibility is that the hub spoke holes and the rim spoke holes are offset by some random angle (not 0 degrees and not 180 divided by the number of spokes). This third option can be thrown out rather quickly - for the hub holes and rim holes to not line up at all, different spokes would be needed for the spokes that pointed different directions. The second option of being misaligned by exactly half a rim hole spacing is actually more feasible, but it turns out that normal bike wheels are indeed very well aligned.

It is therefore important to know that if you extend a theoretical line from the axle to a spoke hole on the rim, that you will also pass through a spoke hole on one side of the hub. This point is not completely obvious since it is difficult to actually line this up visually on an existing bicycle wheel. Knowing this, it is possible to make some very convenient choices for the variables of the distance formula. First of all, let's define (x1, y1, z1) as the coordinate location of a spoke hole on the hub and (x2, y2, z2) as the position of the tip of the threaded end of the spoke inside the rim.

Let's start with z2 as it is the easiest. z2 is simple enough; since the spoke ends all terminate around the lateral center of the wheel:
z2 = 0
If you want to get fussy AND you have rims with offset drilling, it is reasonable to measure the offset and use that value for z2.

Ok, on to x1 because it's really easy. The whole essence of how we're figuring this out relies on:
x1 = 0
In other words, our hub is turned so that the spoke head is pointing straight up or down (it doesn't matter which - let's arbitrarily say down). Since we're controling the hub, let's look at y1 before messing with where the rim hole is. y1 is simply the radial distance from the center of the wheel to the center of a hole on the hub. y1 is the BC radius or half of the BC diameter (BCD).
y1 = 1/2 * BCD

I hope none of that was too confusing because now we get to the more complicated bit. Given our neatly laid out coordinate system, where lies the x2 and y2 point of the spoke end? Here is a mental excercise to help answer that question. Imagine if x2 were 0. It's not, but imagine if it were. You'd have a radially spoked wheel and a lot of this process would be much easier. So you have a mental image of a spoke hanging down from the hub at the six o'clock position. If the spoke was magically the correct but yet unknown length that we are trying to figure out, it would extend well past the rim. If we began to rotate that spoke around the spoke head in the hub hole (keeping everything else fixed), there would be a certain angle where the perfect spoke length sat perfectly in it's correct hole. A good question at this point is which hole is this?

The answer to that question is a bit complicated, but the really complicated part is knowing that this is an important question - and I'm telling you that it is, so that part is taken care of. Using another wheel for reference, you can have a look at the answer to this question. The thing to look for on another wheel or picture of one is the rim interval between neighboring spokes on the same flange that cross each other. In other words, pick a spoke and then pick its neighbor spoke on the hub. If these two spokes don't cross each other, pick the one on the other side of the first spoke. These two spokes should make a wide angle with each other where they cross. The important bit of information is this - on a 36 spoke wheel, there are 9 rim holes between two such spokes. Or put another way, two neighboring spokes that cross near the hub have 1/4 of the rim's spoke holes between them.

That still doesn't answer the question directly, but it is a good start. There are 9 rim holes between the two spokes in question. Forget about all the rim holes that go to the other side of the hub (the other flange). Now there are only 4 significant rim holes between our two important spokes. The important spokes were neighbors on the flange, so their hub holes will correspond radially to the 2 significant rim holes in the middle of the significant set.

That's nice. So what? Well, if we look at our 6 o'clock rim hole (which is significant), we see that it lines up with the 6 o'clock hub hole. The next signficant rim hole lines up with the crossed neighbor hub hole. Keep going. The next one is skipped as it is one of the 4 between our two important spokes, but doesn't correspond with our two important spokes. The next significant rim hole is where this spoke's end goes. To summarize, the rim hole that is radially aligned with a particular spoke head is separated from the ultimate placement of that spoke's end by 2 significant rim holes or 5 rim holes total.

Now we're getting somewhere. If there are 5 holes between the rim hole that we'd use if we wanted a radially spoked wheel and the rim hole we actually are hoping to use, then there are actually 6 spaces between holes. This means on a 36 hole wheel, 60 degrees between our vertical axis and the location of the spoke end.

Ah! Something useful. We don't exactly know where this point is, but we have a much better understanding of it now. The specific x2 and y2 location of the end of the spoke can be calculated using a simple trignometric formula - namely: hypotenuse times sine of an angle = opposite leg of the triangle. Since we based our zero point in the center of the entire wheel, the hypotenuse in this case is simply the rim radius (Rr). The other important bit that we just figured out is that the angle is 60 degrees. So:
x2 = Rr * sin(60)
y2 = Rr * cos(60) = Rr * sin(30)
Spoke Length = sqrt[ ( x2 - x1 )^2 + ( y2 - y1 )^2 + ( z2 - z1 )^2 ]
x1 = 0
y1 = 1/2 * BCD
z1 = flange offset
x2 = Rr * sin(60)
y2 = Rr * sin(30)
z2 = 0
SL= sqrt[ ( Rr * sin(60) )^2 + ( Rr * sin(30) - y1 )^2 + z1^2 ]
SL= sqrt[ ( .75 * Rr^2 ) + ( Rr * .5 - y1 )^2 + z1^2 ] And finally:
SL= sqrt[ Rr^2 - y1*Rr + y1^2 + z1^2 ]

So for a 36 spoke wheel, this is it. It's not really so complex at all in it's final form. The final consideration is some offsets to put the spoke end just where you want it. Up until this point, we have calculated the length of spoke necessary to put the end of the spoke exactly flush with the rim tape (in theoretical-land). This is actually not too useful - especially when you have double-walled recessed spoke wells on your rim. Remember when you had the wire on the rim and I said to measure from the wire to the top of a nipple put into the spoke well? Now is when you can use that number to subtract from the calculated length. This puts your spoke end nice and flush with the end of the nipple where the nipple belongs. We are still in theoretical-land, however and some more adjusting is necessary.

Besides subtracting some length to allow the spoke to sit below the rim tape, there are other things to consider. First of all, the entire calculation measures to the center of the hub hole. Spokes are not easily measured from this point. Spokes are measured at their bend on the head. The distance from this measuring point on the inside of the bend to the middle of the spoke (and therefore middle of the hub hole) is 1mm on a 2mm spoke. This fact tends to give longer spokes than we expect (unless you measure them in an unusual and difficult way).

Next I believe spokes stretch about 1mm on built wheels. They must stretch some and 1mm seems to cover it reasonably well.

Finally, you want to position the end of the spoke just how you like it inside the nipple. My preferences is 2mm below the end of the nipple. This gives some room for spokes that need to stick out a bit for some reason, yet is far enough down that a well trued wheel shouldn't have any sticking out beyond the nipple. If your rim doesn't have recessed spoke wells, then it might be wise to have the spoke tips stop 3 or 4mm below the nipple end to avoid any spokes poking the innertube. This means, however, that you may have exposed threads below the nipple on the inside of the wheel. Besides looking bad, the threads collect grime and are difficult to clean. Perfect wheels don't leave any threads exposed on the spokes either above or below the nipple.

Subtract distance from rim tape to nipple flange
Subtract 1mm for the head offset.
Subtract 1mm for spoke stretch
Subtract 2mm to bury the spoke end inside the nipple

That's it! Now that you know what sizes you need, go buy your spokes. I strongly suggest counting them and then measuring them yourself. Park Tools makes a very easy to use and useful spoke measuring gage, but if you are not that serious about making spoke measurement a regular part of your life, it is reasonable to just mark the necessary lengths (measured from the edge) on a piece of paper. Then lay the spokes on the paper and hook the head around the edge that you measured from. If it lines up with your marks, then it is good. If not, perhaps your bikeshop person is using a "special" way to measure spokes or taking the liberty of "aproximating" a few mm. Spokes are sometimes not available every millimeter, but you should be within 2mm.


Now that you have your rim, hub, and correctly sized spokes, there is a new challenge. "Lacing" is the term that describes putting all of the spokes in the holes that they belong in. There are numerous ways to make mistakes, but hopefully these guidelines will help avoid them.

1. Choose a left and right side for your rim. There are two ways to make this decision. One is by orienting the rim's label to be readable from a certain side. Perhaps you just want the label to be consistent with the other wheel. With many front hubs, the symmetry allows you to spin the wheel around anyway and this isn't much of an issue. If there is some asymmetrical aspect to the hub (a disk brake, or a logo, etc), then this excercise isn't a waste of time. A better reason to make one side of the rim left and the other right is to make the rim's welded seam slide through your brakes in the easiest way possible.

2. Fill one side of the hub's inside spokes. In other words put a spoke into the hub from the outside every other hole. If you have different sized hub flanges (a dish compensated rear hub, for example), do the small flange first since it is the hardest and is easier to do with out any other spokes in the way.

3. A major concept that I use is that inside spokes "trail" and outside ones "lead". This means that as the wheel rotates during normal operation of the bicycle going forward, "leading" spokes hit a certain angle at the rim before they reach that point at the hub. Trailing is the opposite. So if a spoke head on a hub passes under a fender stay and then the rest of the spoke, followed finally by the nipple of that spoke, it is "trailing". There is a specific reason why I use trailing on the inside. First a bike wheel is a tension structure much more so than a compression structure. The torque of the drive axle is transmitted by pulling the rim behind the hub. I feel that this puts more tension on the trailing spokes than the leading ones. I feel that the outside position on the flange is more vulnerable than the inside - especially to overshifted chains. There is a good argument that says that the high tension spokes (trailing) should be on the side that is easiest to replace - the outside. Pick the factor you value most. I'm going to proceed as if inside spokes trail.

4. Once you have figured out which spokes lead and which trail, you should think about where the valve hole is going to be. A poorly built wheel will put the tube valve in the worst, most restrictive place. With some forethought at this stage, you can ensure that the valvestem ends up in the ideal location. Take one of the spokes that you just put in the hub. Run its end to the rim spoke hole that would hit the ground just before the valve stem hole during normal riding. This should put the inside trailing spoke in a position that opens away from the valve hole, not leans in toward it. Think about this and make sure your first spoke will end up leaning away from the valve hole.

5. Offset rim users beware!! At this point, you may or may not have made a serious mistake. Now is the time to check for that possibility. If you have high quality rims, there is a good chance that the rim spoke holes were drilled in a staggered pattern - each hole will be a little off center right then left, right then left, etc. Look closely to make absolutely sure you either don't have this kind of rim or that you have your first spoke going to the correctly offset (same side) hole in the rim. If not, then simply move the initial spoke one rim hole away from the incorrect position in a direction away from the valve hole.

6. Lace the remaining spokes of this inside set. These spokes will occupy every fourth hole on the rim. Just screw the nipple on enough so that it doesn't come off. Leave everything as loose as possible until all the spokes are in place.

7. Now it's time for the inside spokes on the other side. This is a great opportunity to make a very time consuming mistake. By thinking very hard about the problem, it is less likely that you will suffer from it. The problem is if you pick the wrong hub hole for a certain rim hole. It is easier than it seems to make this mistake. First, let's pick the rim hole that will start the other side's inside (trailing) spokes. This one will be right next to the first spoke which was next to the valve hole or one one away from it if you had to move it to match your rim drilling. This spoke is going to go in the unoccupied rim hole closest to the valvestem hole that is on the same side of the valvestem as the first spoke. A bit complicated sounding, but just put the first spoke of the other side in a hole near the valvestem and see if it leans into it. If it does - that's wrong. If it leans away from your valve, that's right.

8. Ok, now that you know where this spoke is going to, where does it come from? The problem is that it comes from the other side and therefore it is not obvious exactly which hole on the other side. However, since you know where the end of this spoke will eventually be situated with respect to your first spoke (near the valve), you should be able to figure out which hole from the hub. The best way to do this is to stick a spare spoke into one of the holes perpendicularly and see where it hits the other flange. If it hits a hole on the other flange, you are doing this wrong. It should hit a space between holes and that will help you line up the left holes to the right ones. A space on the left side of the hub is a hole on the right. Since your first other side spoke is next your very first one, it would be between holes if it was on the same flange as the first. Use the alignment spoke to see which hole corresponds to that space on the first side.

9. Lace the final inside spokes. The pattern on the rim should be 2 holes full, 2 empty, 2 full, 2 empty at this point.

10. If there's a choice, pick the small flange. Stick a spoke from the inside of the hub so it becomes an outside spoke. This spoke will be a leading spoke. Angle it away from the others on the inside. So where does this spoke go? Between this spoke and the same flange side neighbor which it immediately crosses, there will be 9 rim holes. I don't mean the 9th hole; I mean that there will be 9 other spokes in positions along the rim between any given spoke and its nearest crossed neighbor spoke on the same side of the hub.

11. So now that you have found where this spoke is supposed to go, there is another important detail - crossing. To get this right, temporarily forget that the flange you're not working on exists. Concentrate on just the spokes on the side you're working on. These leading spokes "cross" the trailing ones. The normal type of wheel and the one I'm describing here is called a "3-cross" pattern and this is why. Each spoke forms intersections (as seen from the side) with 3 other spokes from the same flange. The first is very close to the hub as it crosses the neighbor spoke right away. The next one is crossed a bit farther out - it's hole is 60 degrees or 3 hub spaces away. The final one is the important one. It is 100 degrees or 5 hub spaces away. It is important because the outside spokes that you're working on must pass this third spoke on the inside. This makes them actually touch at this point whereas the other intersections are visual only. It may require some slight gradual bending to get these spokes in position behind the others. Once you've got the crossing pattern figured out, fill the rest of that flange and weave those spokes into position.

12. Now you have 9 empty hub holes and 9 empty rim holes and an imbalance of trailing spokes. Fill the remiaining holes. At this point everything should make sense.

13. Finished! Things to check - Are all of the spokes physically touching exactly one other? If not, you didn't cross correctly. Does the valve stem hole show up between the widest spread apart spokes giving the best access to it? If not, you chose your first rim holes poorly. As you drive the wheel forward, will the brakes see the easy direction of the weld seam? If not, you picked the right and left of your rim incorrectly. Are the spokes from the right flange going to right biased holes on the rim? If not, then you forgot to check this when you chose your first spoke hole on the rim.


At this signficant point, you should have something that looks almost exactly like a bicycle wheel. If you followed instructions so far, you will have laced the spokes so that the nipples are engaged on them to the minimum degree necessary to keep them attached. If the spoke sizing exercise went well, this should yeild an extremely wobbly and unsafe wheel. This brings us to the next phase: tensioning and truing.

Let's talk about the rims to introduce my idea of wheel tensioning. If you purchased a good rim, things will go much better for you at this stage. The criteria for classifying a rim as good in this context is how well did it lay on a flat surface before you laced it. If it was possible to lay the rim on a glass table (or precision granite surface plate) and there weren't any nasty gaps along the edge, you have a good rim. This will greatly facilitate truing. The reason is that if you have a rim that started out with some lateral misalignment, you will need to have spokes that have more drastically varying spoke tensions in order to compensate. Whether you have a good rim or a bad one, I would recommend proceeding the same way, so don't go to lentghs to try and measure your rim's defects. A flat rim will just make things easier.

Before you go attacking the wheel, here are some notes to make this operation easier. First, you can use a screw driver to adjust the spokes from the outside of the rim, but I strongly recommend a spokewrench (or as the Germans more correctly call it - a nipplewrench). The black Park spokewrench works for DTSwiss and SAP brand spokes. Second, keep in mind that the valve hole is a good indicator of a position on the wheel. Start there and end there when doing things uniformly to the wheel. Keep in mind that the rim label is usually half a wheel away from the valve hole and this fact contributes to the fact that a wheel left hanging will usually fall so that the label is down and the weightless hole is up. Third, you need to turn the spokes the correct way. It doesn't matter how great of a mechanic you are - when playing with spokes poking every which way, there will appear with surprising frequency opportunities to be unsure of which way to turn the spoke to meet your objective. Think of it like this: to tighten a spoke, the nipple will move closer to the hub; to loosen, it will move away. That fundamental property is easier to visualize on the other side of the wheel than specifically which way to turn the nipple. To find out which specific way to turn the spoke nipple, point your right thumb in the direction you want the nipple to go, either toward the hub or away from it. If your thumb is pointing in the desired direction, your curved fingers will show the direction of rotation necessary for this.

The first thing I do is to tighten the nipples without a wrench until there are only about 1mm of threads left showing. This just means going around to each spoke in turn and taking up most of the slack. If you find that it is getting hard to turn the nipples, then maybe 1mm isn't enough. Basically, you want the nipples to all to be finger tight to the same position on the spoke. The position can be observed by the spoke threads.

Next I would move the nipples until all of the threads were gone, but only just until this point. Use the wrench if necessary, but if it is really hard (it should be relatively easy at this point), this may be too far for this stage. By doing this, I know that all of the spokes are effectively operating at the same length at this point. Since our rim and hub are high quality, this should, in theory be setting our hub up right in the middle of the wheel.

Knowing that I have all of the spokes engaged to the same extent is a nice way to start out. Once the threads of the spoke disappear into the nipple, it won't be possible to use them as an indicator of this. From this point, I use the spoke wrench in a more controlled way. So ideally, when the threads are just getting ready to disappear into the nipple, the wheel is just about to take on the right tension. Go around the wheel and tighten each spoke until it is just on the loose side of correct. Turn each nipple almost exactly the same amount to preserve the relatively even engagement that you obtained visually using the threads. I say almost because I find it convenient to make sure that the wrench flats on the nipples are all lined up in the same way. This makes it easier to quickly get the spoke wrench in and out. By lining up each nipple's flat surfaces, you can only screw things up by a maximum of 45 degrees per nipple.

You're probably wondering what I just meant by correct with respect to the final spoke tension. This is a good question and neither easy to explain nor know. The best thing I think you can do is to be experienced at knowing what correct wheel tension feels like. Perhaps the best way to do this is to feel the spokes on existing wheels. Take a trip to a bike shop and play with spokes on various wheels until you feel like you have some understanding of the right tension. Keep in mind that you are balancing two conflicting ideas. A wheel that is too tight will cause obvious tension in the spokes and will lead to their premature failure. A wheel that is too loose will cause too much flexing of the wheel components and maybe even a wobbly ride. Besides the wobbly ride, the wheel will have a tendency to come out of true and you will experience greater spoke breakage from all the flexing.

Now you have the wheel slightly loose and you know that the nipples sit on the spokes in just about the same position uniformly. This is the best situation to start really truing the wheel. At this point, you need a truing stand. Well, a truing stand of some sort which doesn't autmatically mean an expensive tool dedicated for the job (though I'm sure that would be good too). For a truing stand, I use my bike's frame. In fact, by using the frame of the bike you eventually want the wheels on, you will end up with a less perfect wheel, but a more perfect wheel for that bike. To illustrate this idea, imagine that I create a wheel that is perfectly aligned with a poorly built, flawed frame (imagine one brake post welded at the wrong angle). The wheel will also be flawed. However, it will be perfectly aligned to said frame - better, in fact, than a pefect wheel.

So spin your bike upside down and figure out a way to support it. Stick the wheel in question into the rear dropouts or forks depending on which it is. Spin the wheel and have a look at how things stand to start with. Is the wheel really out of true? Hopefully it is not horrible (though you may need to open up your brakes to start with). Usually at this point, the wheel is within 3mm without having done really anything. What I look for at this point is general alignment over the center. What I do is measure from a right flange spoke at the nipple to the center of the right brake mounting bolt. Then I measure from the left brake mounting bolt to a left leaning spoke in the same way. Take a few samples like this around the wheel and you can get a general idea of whether the rim is too far to the left or right. If so, and this is especially probable with the back wheel which wants to be dished, I tighten by a fixed amount (one turn maybe) all of the spokes on the side that is too far away from the rim.

Hopefully, the rim should be roughly centered between your brake posts. Now you need to get things lined up better. The absolute best way to do this is to take a machinists dial indicator and fix it to a brake post. With this method, you can get extremely true wheels with a minimum of fuss. These dial indicators are not all that expensive and definitely make things easier and better. If you don't have a dial indicator, you can use the brakepads as an indicator. Afterall, they are going to be one of the more important reasons to have a true wheel in the first place.

However you are keeping an eye on the lateral position of the rim, the general idea is to find all the spots that are too close and all the spots that are too far away. If a section of rim is too far to the left, tighten the spokes in that region that go to the right. If you feel that there is too much tension in this area already, try loosinging the spokes that go to the left. Don't make big adjustments. Start with half turn adjustments in regions of 6 or so spokes. As the tension and trueness increase, drop down to quarter turns.

Once you have the wheel spinning fairly well, under 1mm of lateral motion, it is important to make sure some other things haven't become screwed up on the way. First is to check to see if you have inadvertently pulled the rim out of a general centered position. Is the rim closer to one brake post or another using the technique to measure that described above?

The other thing to watch out for at this point is eccentricity. Obviously you're somewhat of an eccentric person if you're building your own wheels, but I'm talking about wheel eccentricity or ovalness. I lay a straight edge across my chainstays right up against the two flanges of the rim. Spin the wheel. The highest spot of the wheel will push the straight edge away from the center. Note the distance between it and the rim at the lowest positions. This is your eccentricity error. Don't get too excited about making this perfect. Sometimes it just isn't possible. I tolerate a couple of millimeters if the error is gradual. Any more than that, tighten all the spokes (left and right) in a high region and loosen all in a low region and see what you get.

Once you are happy with the eccentricity and the centeredness, go back and fine tune the lateral trueness. At this point, you should be making small adjustments of a quarter turn for only a few spokes at a time. Also note any spokes that feel really tight which are already hard to turn. Leave these alone if possible and work with the other side to acheive your desired effect.

Eventually, by making problems around the wheel smaller and smaller, you should get to the point where the error is not worth playing with any longer (using an indicator, I shoot for an error less than 0,1mm [.004"]). Everything should almost be ready to go. There are two things I do after truing the wheel that I feel contribute to a better wheel. Both involve reducing stress concentrations.

First, I stress relieve the spokes in what may seem like a crude way, but it is very effective. Grab a group of 4 left spokes where they form two nearby crosses. Squeeze this group together nice and hard. You can do this on both sides at the same time with both hands. Do this nine times or until you have pulled hard on all the spokes. This will help shake out any torsion that the spoke threads have imparted to the spoke and seat the spoke heads in the hub and the nipples in the rim. After some creative pushing and pulling of the spokes, things should better reflect the condition that the wheel will be in after a few km of riding. Better to find out if things settle now than later.

The other thing I do is a bit more elegant, but not so easy. I take a metal thing - almost any metal thing will do; I use a metal 150mm machinists scale. Use this metal thing to lightly tap on each spoke. The sound that this makes can be used to determine (roughly) relative spoke tensions. Specifically, what we are looking for with this method is one or two spokes that are totally out touch with the rest of the wheel. This seems like it would be next to impossible if you aren't an accomplished musician with perfect pitch, but it's not really so difficult. Start with the realization that the spokes on the inside of the hub flange are going to be (generally) at a slightly different tension (less) than the ones on the outside. That means that as you tap spokes from one side of the wheel, you'll get a pitch pattern of high/low/high/low, etc. Listen for this. Start humming the pitch interval. Any spoke that is drastically not tensioned like the rest of the wheel will be rather noticeable with this system. If the pitch is too high, loosen the spoke a bit - too low, tighten it.

After doing all this, check to see if the wheels are still okay with respect to eccentricity and that they are still laterally true. If not, go back and play with the problem areas some more. If they are, you are finished! You should now be able to enjoy the pride, confidence and performance of self-built wheels.

Return to Bike Page
This page was created with only free, open-source, publicly licensed software.
This page was designed to be viewed with any browser on any system.
Chris X. Edwards ~ 16 April 2001