Disclaimer: this has been asked and nearly answered, but I'm still missing some detail.

I have a symmetrical rim with offset spoke holes (Spank Spoon 32, 26", 32 holes, see pic) which I want to lace to my existing DT Swiss rear hub. There are two ways of lacing this (given the usual 3-cross lacing pattern):

Version 1 (increased bracing angle) - drive-side spokes laced to left-side holes, non-drive-side spokes laced to right-side holes. Increased spoke bending near rim, but laterally stiffer wheel.

Version 2 (reduced bracing angle) - DS spokes are laced to right-side holes, NDS spokes laced to left-side holes. Less spoke bending near the rim, at the cost of a laterally weaker wheel.

I believe version 2 is the default, because manufacturers want to avoid the common failure mode of spokes breaking just below the nipple. However, Sapim manufactures nipples with a hemispherical head which should alleviate this issue (by enabling the whole nipple to bend in line with the rest of the spoke), so long as there is enough tolerance between the nipple and the rim hole for the nipple to actually bend as intended.

So: I'm inclined towards trying "version 1", although it's unconventional. Does anybody have experience with this? Is it a bad idea?

Spank Spoon 32 rim

2 Answers 2


Version 2 is what is usually done. If you do version 1, please ensure the nipple can actually swivel enough in the spoke holes. The thickness of aluminum in the rim and the spoke hole size sets a maximum swivel angle for nipples. If the spoke needs to swivel more, it may run to a hard stop and instead the bending may happen not at the nipple but along the spoke. If this is the case, you want to correct the spoke line such that the bending at the spoke is not a smooth gradual bend, but a hard abrupt bend. See "correcting the spoke line" in The Bicycle Wheel by Jobst Brandt.

I do all my wheelbuilds on hemispherical nipples (not Sapim but DT Swiss Pro Head). These are the only nipples that should ever be used. Even for offset hole rims built the normal way, or for non-offset hole rims, all you should use are the hemispherical nipples.

Note that it's not enough to just have hemispherical head nipples. You also need the spoke line at the hub flanges to be good. Some hubs may have the hub flange slightly bent, to assure good spoke line for spokes on both sides. This slightly bent hub flange is intended for rims with non-offset spoke holes, or for rims with slightly offset spoke holes where the wheel is built the traditional way. Your way (version 1) might not be what the hub manufacturer thought by bending the flange slightly.

If you haven't purchased the spokes yet, prefer triple-butted (2.34mm / 1.8mm / 2.0mm) spokes. Butted spokes, thinner in middle, are better than non-butted spokes. The triple butting is required because optimally spoke hole at the hub would be just big enough to pass a spoke through. Unfortunately, it's slow to lace a wheel with correct spoke hole size at the hub. In order to increase wheel building productivity, the hub manufacturers have made the spoke hole overly large. This is a problem because 2.0mm spokes have rolled not cut threads, so the thread is slightly larger than 2.0mm. The mismatch between the spoke hole in the hub where it's very fast to push the over-2.0mm threads through, and the exactly 2.0mm spoke J-bend end, means the ordinary 2.0mm spoke doesn't have a good match to the oversized hub hole.

The triple butted spokes (2.34mm / 1.8mm / 2.0mm) have optimal J-bend size, meaning the interface between spoke J-bend and hub spoke hole is perfect. It's slightly slower to lace a wheel with triple-butted spokes but the difference in time is paid back by a more reliable wheel.

  • +1 for the reference to Brandt's book, great tip. I should have mentioned that my hub uses straight-pull spokes, so the discussion around J-bends doesn't apply to me. (Although the general advice to use butted spokes is correct, as elaborated on by Brandt.) Great answer, thanks.
    – zire
    May 28, 2022 at 17:32

Some manufacturers have produced rims and wheels that increase bracing angle on purpose in exactly the same way you're proposing, even though it's the unusual way in the scheme of things. Shimano wheels circa 1997-2002ish for example featured an extreme version of the concept. Some BMX companies have also done it.

You could build it the way you propose in #1 and it will work fine as long as the nipple can freely swivel enough to be aimed at the hub flange. If it only barely or almost does, that will increase friction of the nipples turning under tension and generally make the wheel a mess to build, even if you carefully set the line at the nipple later in the process.

One problem with doing this is that whether or not the nipple swivels that much is a function of the rim's lower wall thickness and how the drilling is done. If you did this and then ever needed to rebuild it with a rim that didn't accommodate it in the same way, that would then mean double-grooving the hub, which introduces a risk of flange failure and is bad form.

I think doing it as a fun hacky experiment is reasonable, but if the hub you're working with is any kind of high end, indefinite service life one, it's probably not a good idea.

  • Your answer as well as juhist's are fairly similar, but both excellent. My core takeaway is that the nipple needs enough room to swivel in the rim, which I'll be careful to verify. Thank you.
    – zire
    May 28, 2022 at 17:23

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.