Pictured below are two wheelsets built with 650b Pacenti Brevet rims, one built in March and another more recently. I was happy with the steadiness of the rims and would recommend them if they meet your needs. The first set uses Campagnolo hubs; SON and Shimano hubs on the second. The hubs were supplied by their owners except for the SON dynamo, which I provided. The Rene Herse (formerly Compass) tires were provided as well, an option for wheelbuilding customers.
Rear hex drive nipples were designed to handle increased torque but they also prevent builder’s marks on external nipple wrench flats. Some people like rear drive because the turning direction is intuitive — clockwise is tighter. Since the external interface exists just the same, you can forget the rear hex drive if you prefer (or treat it as a backup to save the day if a nipple cracks or rounds off).
The Sapim hex drive is 5.5mm. It’s a balance between making the interface as large as possible while leaving room for tool clearance at the rim holes. Even so a thinwall tool is often necessary. I have three in my toolbox. My main tool is the Park SW-15, a 3-way that fits nicely in the hand and provides good leverage. Blue tape marks the 5.5mm end. The Park Tool SW-18 is a screwdriver design, which reaches into deeper rims and may fit better in race toolboxes. The Sapim factory tool is excellent quality and handles even deeper rims but may be too deep for everyday use.
This is a sweet set of 650b gravel wheels. They’re put together using Onyx rear and SON front hubs paired with Light-Bicycle carbon rims. These wheels are destined for a year-long journey so I’ve selected Sapim Force spokes, which makes builds more bulletproof. For more casual use lighter spokes would definitely work. Please email for help sourcing SON generator hubs.
The correct spoke length is one that penetrates the nipple just right — not below the screwdriver flats and not past the top of the nipple. You don’t want to go too short because spoke threads reinforce nipples internally. Nipples, particularly aluminum nipples, have higher failure rates when paired with short spokes. You don’t want to go too long either. I define too long as past the end of the nipple. This is strictly true for some nipples however the nipples sold here can tolerate a couple turns past the end without issue. Turning past this baked-in limit causes nipple threads to grind into the unthreaded spoke shank, which creates a weakness in the spoke and is associated with spoke breakage.
The rounding part
Spoke length calculators output lengths to the tenth of a millimetre, which must be rounded to match available supply. There’s no universal rounding algorithm because it depends on how you calculate spoke lengths in the first place. It depends on ERD measurement technique.
In my process I use measuring tools based on theoretical spoke penetration to the bottom of the nipple screwdriver slot. I don’t want my actual spoke penetration to be short of this mark, so I’m cautious about rounding down. My baseline is rounding to the nearest millimetre. If my target tension for a spoke is high, say 125kgf / 1200N, I’m not bothered rounding down as much as 0.6mm. This is because spokes stretch a fraction under tension and stretch is more pronounced at high tension. I’m not bothered rounding down 0.7mm if the spokes I take to high tension have an extra-skinny diameter (the minor diameter in the case of butted spokes). Thinner spokes are more elastic. The opposite holds as well — a spoke built to 60kgf could be rounded up 0.6mm. As an example when building a rim brake rear wheel, where tension is quite imbalanced between sides, a wheelbuilder might round up on the low tension side and down on the high tension side if both lengths are halfway between sizes.
(Hopefully you can see this rounding logic doesn’t make sense if your theoretical spoke length targets the end of the nipple. In that case you would rarely round up.)
Traditionally spokes are stocked in two millimetre increments, which invites compromises depending on your calculated lengths. You can be forced to reckon with the the risks of going too short or too long. Maybe the uncertainty steers you to brass nipples instead of aluminum. Maybe it causes you to pick up nipple washers as an insurance policy. Bottom line: it’s not doing you any favours and to build the best wheels you ought to shop in one millimetre increments.
About this series
This is a series about spoke length. These aren’t tutorial articles but rather a collection of practical hints. The focus is correct inputs to spoke length calculations — correct input leads to correct output — and how to use outputs. For more general information, see my books list.