Got a pair of 32-hole Ambrosio Excellight rims the other day, and began the process of lacing them up to Record hubs. According to the spoke calculator in Excel format by Damon Rinard, for a 3-cross pattern the spokes for the front wheel should be 294, while for the rear they should be 294 and 292.

The rear wheel worked out perfectly. But not the front: for some reason the last four spokes come up short by a few mm, and simply cannot be attached to the nipples. I checked them for length, and found them to be 294 mm as claimed. I even took a few of the other spokes off and checked them for length--no discrepancy there, either.

Is this simply a matter of needing 296 mm spokes for this particular wheel, or am I missing something?

Without having the opportunity to see your wheel my guess is that you have an error in the laceing pattern. The most common is to have the right side spokes 1 hole off from the left side spokes. If, when looking at the spokes that are already laced, you see 2 tight spokes followed by 2 loose spokes, that's it.

Well, there was nothing wrong with the spoke pattern. The rim isn't out of radial true much, but lateral true is out. Somehow it seeems that the hub is not exactly in the center of the wheel, but is a few mm away from center, and this is causing the last four spokes to fail to reach the nipples.

Do I need to twist the hub, maybe? I'm flummoxed.

Start over. Is the rim round or does it have flat spots? If the rim is good then as you install the nipples turn each of them onto the spokes 2 turns only. It is necessary to twist the hub a good amount when adding the last two sets of spokes. Once you have all of the spokes installed check the pattern. As you bring the tension up turn all of the nipples the same amount. Check the radial true as you increase tension. It is possible to have even tension on all spokes with the rim out-of-round. It's best to concentrate on radial true during the earlier phases of the build.

Did you forget to make sure that all currently installed nipples have been pulled into the rim holes - i.e., there isn't one or two jammed up against the inside of the rim preventing you from getting the last few spokes to reach their nipples and holes?

=8-)

Yeah, I've tried everything, but I'm still scratching my head.

How about some pictures?? The only other thing I can think of that has not been mentioned should be obvious - spokes are in the correct offset on the rim?

I'm pretty sure that there is.

Each side can be correct within itself but, if the two sides are off in their relationship to one another by one hole, it'll do what you got. Don't ask me how I know.

I just went and tore the thing apart. I'm going to start afresh tomorrow.

This is the final thing that occurred to me. According to Roger Musson's excellent wheelbuilding book, you do the first eight spokes on one side of the hub, and then the first spoke on the other side goes through the hole which is offset just slightly to the left of the spoke you started with when looking at the hub from the side. Maybe I didn't get that right? I'll see how it goes tomorrow.

Light just went on: I have a Campy Record rear hub that a friend gave me in a box of old parts in which the holes on the two flanges are not offset from each other. They are exactly opposite one another: a spoke passed perpendicularly through one hole goes right through the hole on the other side. This caused problems when I tried to lace it up with new spokes: some are either too short or too long. My friend didn't build his own wheels when he was racing way back when and he doesn't remember anything weird about wheels built on that hub. But note that by the time he gave it to me it was just a hub, no longer a wheel. So perhaps there was a "story" there.....

If you look at your hub and discover that the holes on the other side aren't offset, Bingo!

I would think that the only way you could get this to work would be to figure out how much longer (or shorter, depending on which side you did first) half the spokes would need to be in order to reach the rim from a point on the hub that is 1/32nd of the circumference farther (or closer) than it should be.

Yup. That's my bet. I've done it more than once myself. Very frustrating.

My trick is to stick a spoke through a vacant hole and line it up as parallel as I can eyeball with the hub body. It'll hit the gap between 2 spoke holes on the opposite flange. Then I figure out which hole I need to start laceing the opposite side.

I will admit it's been a while since I built a wheel, but I built hundreds as well as wrote up instructions as part of a mechanic's course I designed. I don't think that method works on all rims unless things are more standardized than in the past. When I was actively working some hubs had different offsets next to the valve hole (right side vs. left side offset toward you) and the spokes had to be offset differently on the hub if you wanted to have the valve stem open (between sets of crossed spokes).

Bingo! Shazam! Pepsi-cola! Home run!

This is the problem exactly. You're a genius.

I just redid the wheel to great success, and have got an even better fix: after doing the first eight spokes on the driveside I poked the first spoke on the non-driveside through the hole just to the left of the one directly opposite, and everything laced up perfectly.

The problem was not the hub, as you just stated the holes are offset - you just went in the wrong direction previously when you went from one flange to the other. (right instead of left). That made the lacing look right but each spoke had to reach further than it should have, pulling the hub off center until the last few would not reach at all.

To be precise, as per conspiratemus1's post above, the holes seem to be lined up more or less perfectly across from each other; they're not offset. In my first attempt I started the opposite spokes directly across from the ones on the first side, but that didn't work. What worked was putting the spokes on the opposite flange one to the left.

Sweet. Glad to see it can be done. Just to be certain, I take it you were able to use the same length spokes for the two sides -- 294mm --. correct? Do the spokes all engage the nipples to the same depth, or is there an alternating high-low pattern, even if just a millimetre or two, as you go around the rim?

When I build a wheel, I start with all the spokes showing two threads exposed, then go around the wheel tightening a turn at a time until tension starts to come up. This is usually when the spoke end starts to eject the little spur on the end of my nipple driver or when the nipple driver can no longer exert enough torque to turn the tightening nipple. If the spoke holes are not offset, it would seem that the spokes on the second side are going to engage a little farther (and eject the nipple driver sooner), since they start 1/16 of a flange circumference to the left but end in a hole that is only 1/32 of a rim circumference to the left. I suppose the act of centering the rim on the hub will take care of this -- just make sure you do have spoke tension as close to equal as you can get it.

I don't understand why half the spokes should be longer, hole offset or no.

Surely it's a simple case of splitting the difference; the flange holes are then 1/64 away from their normal locations for each side.

You might well be right. I'm hoping Rousseau will tell us, since he's built the wheel.

But consider this: imagine a hub with only one flange. If you lace it up with pulling and pushing spokes all brought to the same engagement, the rim is then already fixed in position by that single flange. Now if you imagine the second flange added with the spoke holes exactly opposite the first flange, the rim can not shift that 1/64 revolution "around" the hub to split the difference because the spokes in the first flange have immobilized it. The second set of spokes will be too short or too long. That is, after all, why Rousseau's first attempt at the wheel failed. (I know you don't normally build wheels one flange at time, but this thought experiment explains why both sides can be correct individually but wrong when considered together.)

I'm sure I'm right; the wheel is symmetrical.

I seem to recall in Musson's book that he has never successfully laced a wheel with non-offsetting flange holes, even with mix-matched spoke lengths. It's been a while since I read the book though... and hey, you got it to work anyway so congrats.

You can't be sure you're right about anything until you've tested it. Find a hub with non-offset spoke holes, build a rim on it, and let us know how it comes out.

Uh, no - given unambiguous and correct information, some things can be said with certainty. This is one of them.

There's nothing asymmetrical about the wheel, so it should build up symmetrically.

After reading your edit (which seems the right explanation for the issue of the OP), I'll add the qualifier that of course the wheel needs to be built properly...

IMO you should have enough spoke length to fully lace a rim loosely enough in order to avoid such hassles. In this case the calculated length seems marginal given the lack of offset flange drillings... could happen to anyone the first time.

Never come across one, myself

Forgot about this thread. I've been enjoying riding on my newly-built wheels too much!

Yes, I used the 294 mm spokes. And they did all engage the nipples to the same depth.

Thanks for the tip, I'll try that on my next build (whenever that is). What I did on the three wheelsets I built so far is twist the nipples right to the edge of the thread. Seems to work, but I'll try your way.

Interesting. These are 1998 Campagnolo Record hubs, so they can't be that rare, can they? Of course, I have no way of precisely measuring whether the holes on both flanges are exactly lined up across from each other or not, so maybe those in the know wouldn't consider them to be as such.