My first wheelset build here. I am following Sheldon and this guy. They both stress very much to tighten the nipples to an equal length, for example just as the spoke threads stop showing, and the even more carefully, do equal number of turns on each nipple, in small increments.

Building the front wheel I followed this advice and the result was several very tight spokes and a wheel out of true. My next step was basically to make the tension on all spokes equal, and then true the wheel.

The wheel is true now and looks ready for use (dishing error, vertical and horizontal mis-alignment are small).

So why equal number of turns, if in the end equal tension is the desired result? Or did I do something wrong?

On another note, the rim freshly-purchased, should be true (looks true, and carrying the two rims next to each other, there were no gaps between them, meaning at least sideways being true). So it makes sense to tighten the spokes with equal tension, to avoid twisting the rim, right?

PS: I triple checked that the nipples are well seated into the eyelets and that the lacing pattern is equal everywhere and the same as on an another wheel.

  • 2
    Counting turns does not take into account possible diferencies among spokes or among nipples, something very likely on low end ones. Another factor may be te fact that some spokes are head-in and others head-out at the hub. Also, this doesn't take into account dishing like for a front disk brake hub or a rear hub for a high cog-count cassette, SO I may say that couting turns is not a rule but a guide line to apply at the beginning when there are still spokes missing and before any of them gets tightened. I'm no expert thoug, I have built 2 wheelsets from scratch.
    – Jahaziel
    Sep 23, 2014 at 21:40
  • 3
    You start out with the assumption that the rim is round and the spokes are all the same length, and you strive to get all the spokes to near the same tension based on those assumptions. But at some point in the truing process you must deal with reality and start adjusting the spokes individually (or in small sets). Sep 23, 2014 at 22:54
  • @DanielRHicks, the spokes deviate by sometimes as much as +-1mm from the calculated length, because no one sells 255.7mm spokes.
    – Vorac
    Sep 24, 2014 at 6:56
  • @Jahaziel, furthermore, as per Sheldon, one can hammer the leading spokes to bend them at the hub, bit cannot access the trailing spokes on the inside of the wheel. I am confused because both sources (Especially Mike) stress the same number of turns very much. Do a Ctrl+F on his site for "go round 5x doing half turns".
    – Vorac
    Sep 24, 2014 at 6:59
  • But is does NOT instruct you to "go round 5x doing half turns". It says "It's far better to go round 5x doing half turns than to go round once doing two turns and then find out you can't get two turns out of the last few spokes." "Go around once. If they are still mostly loose, go around again with another turn." If you kept going on that step after you had several very tight spokes then you did not follow the instructions.
    – paparazzo
    Sep 24, 2014 at 13:14

2 Answers 2


Daniel and Jahaziel are bang on with their comments, but it is probably worth expanding a little.

Taking the front wheel, as this is more straightforward:

If you assume a perfect world, a round rim, correct length spokes etc, then if the tension in every spoke is the same, and the tension in every spoke is the correct value for the rim, then you will have a perfectly true wheel.

You can say this because a front wheel is built symmetrically. For a rear wheel (assuming a cassette or freewheel comes into play), you need to consider drive- and non-drive- sides separately. Anyway...

Obviously since the tension in the spokes is governed by the number of turns of a nipple key, you can under these circumstances say that the number of turns should be consistent across all spokes. So this is where the theory is coming from.

The problem comes because of imperfections. The rim may not be perfectly round, and/or the wheel as a whole might track to the left or right, rather than straight ahead.

To correct these imperfections, you need to adjust the tension in the relevant spokes (or, again as Daniel says, groups of spokes) to bring the wheel into true. So you will be tweaking spokes' tensions such that they likely end up slightly different to each other. This is where practise deviates from theory. So far from it being "important to keep the number of turns the same", it is actually pretty inevitable that they will be different.

As a couple of general comments, first of all, I was taught to build wheels by a real person (who was themselves a good wheelbuilder) - I would hate to have to rely on Sheldon or YouTube for this, so.....respect! What you're doing is not necessarily easy. Second, as a guide, getting the wheel "roughly" correct is quite straightforward, getting it perfectly true is somewhat more time-consuming. You'd probably expect this - there is definitely an 80/20 thing going on here.

  • So, in theory, if the rim I start from is wrapped, but the spokes are the correct length, following the rule of equal turns should actually straighten up the rim (in the perfect case making subsequent truing of the wheel unnecessary)?
    – Vorac
    Sep 24, 2014 at 13:35
  • sorry @Vorac I don't understand that question...wrapped? Do you mean warped?
    – PeteH
    Sep 24, 2014 at 14:13
  • 1
    Indeed. Whatever, this seems to be a complex issue, not easily solved by someone on the internet. Probably because the sensor (me) is a source of noisy and incorrect information. I'll post if I learn more about this discrepancy.
    – Vorac
    Sep 24, 2014 at 14:41
  • 2
    No, if the rim is warped, you'd expect quite different tensions/number of turns. Some spokes will be carrying only the "base" tension, and some will be doing extra work correcting the rim.
    – Useless
    Sep 24, 2014 at 14:51

Then you did not follow the advice if you had several very tight spokes.

The important thing at this stage is to get all 36 spokes to be as close as possible to the same setting, all pretty loose.

The idea is a starting point with equal lengths. And all pretty loose.

On your wheel threads out of site was probably too much of a starting point. Once the treads go out of site the only reference you have to one relative to the other is to count turns.

  • I did 1) tighten all spokes to within 4 threads showing then 2) tighten all spokes so that threads just disappear then 3) go round the wheel two times doing half a turn on each spoke. The result was some tight spokes, and some wobbling. Both wheels are dished.
    – Vorac
    Sep 24, 2014 at 7:02
  • 1
    @Vorac - Note that the spokes may not all have been threaded for the exact same distance. This is especially true if the spokes come from two different batches. Sep 24, 2014 at 12:02
  • If you had some tight spokes then you did not follow the instructions. "All pretty loose" is an instruction. That instruction overrides same number of turns.
    – paparazzo
    Sep 24, 2014 at 12:47
  • I would start even with 2 rounds each spoke, if more will be too tight. Then you should look at your rim every spoke turn, to check it trueness.
    – Alexander
    Sep 24, 2014 at 12:49
  • @Blam, a sane conjecture, but it conflicts the "thread all spokes until just no more thread is visible, from there on you have no frame of reference, then do equal number of turns". I think I will follow it for the rear wheel!
    – Vorac
    Sep 24, 2014 at 13:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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