I recently built up my first set of straightpull wheels, with I9 hubs and traditional DT Swiss Competition spokes. (Hubs were from a I9 Trail S Hydra wheelset and are very much like DTs.)

Something I didn't expect, but ran into, is that when finishing up tensioning the wheels, it seemed like the inner and outer spokes on each side required a slightly different tension; about 1 number on a Park TM-1 tension meter.

This seems to makes sense because the spoke holes for inner vs. outer are a few millimeters different when it comes to flange distance, and thus the inner (steeper bracing angle) spokes would likely have a slightly higher tension.

The wheels built up wonderfully and I had no problem getting to even tension, but it was a bit surprising at first because instead of having one final tension for each side of the wheel I had a (slightly) different for inner and outer spokes on each side of the wheel. For example, here's the numbers from the front wheel:

  • NDS (L) Inner: 22 (117 kgf)
  • NDS (L) Outer: 21 (105 kgf)
  • DS (R) Inner: 20 (94 kgf)
  • DS (R) Outer: 19 (85 kgf)

So my question: Is this normal?

Due to what's effectively a slightly different flange spacing it makes sense, but I wasn't able to find anything about it searching around online, so maybe it's just one of those things that's understood/accepted by more experienced wheelbuilders?

(If it's useful I have a full writeup on that build here.)

  • The spokes cross each other, so they pretty much have the same angle. The difference is only the thickness of a spoke. – Michael Feb 27 at 19:30
  • Why does it seem that the spokes require different tension? – ojs Feb 28 at 17:52
  • ojs: If I'd take, say, the NDS Outer spokes at 21 and try to bring them up to 22, the Inner spokes would drop from 22 to something lower. At first I thought this was a spoke bed thickness issue, but it was consistent between inner and outer spokes. – Steve Vigneau Mar 1 at 22:26

Yes, it's normal and you correctly understand why. Some straight pull hubs can create more variance in bracing angle per side than a j-bend hub would. Other straight-pull designs (Velomax/Easton comes to mind) put the spokes in a straight line, creating the least of this variance possible.

On a j-bend hub, there is also some amount of this effect. But, it's usually very minor. How minor varies with flange thickness and rim diameter. In other words, it's at its least in a road wheel (big rim, dainty flanges) and greatest within BMX (little rim, sometimes thick flanges). You see the same thing where it's almost never a noticeable phenomenon building road wheels that the true optimal spoke length varies with the side of the flange, but if you look at nipple thread engagement on most BMX wheels the difference in engagement is readily observable. In almost any case with a j-bend hub, the effect is minimal enough and the rim is stiff enough that you can have all the spokes at the same tension and the wheel will still be visually true, because the rim stiffness is enough to make it that way. But, using a dial indicator on a very light, low spoke count rim built to near-perfect tension tolerance, you will be able to see the effect on the indicator.

Another way to think about it: the classical spoke length formula reduces the wheel to a bunch of circles and triangles between points in 3-dimensional space. It doesn't account for flange thickness unless you run the formula once for each side of the flange. It also doesn't account for the thickness of the spoke, i.e. the distance between its surface and center.

In practical terms, for bikes, once in a very great while, if you build a lot of BMX or ebike wheels, the whole thing can result in running out of threaded length on one side of a flange but not the other. I've seen it a few times. The importance of understanding this effect is actually in resurgence, because small wheel motor hub rebuilds are an increasingly common shop task, and they have some of the thickest, shortest spokes on the thickest flanges of pretty much any bike wheel genre in history.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.