I had the opposite problem with my first wheel build. The spoke lengths were ok for a 12mm nipple, but too long for a 14mm nipple. I initially laced with 14mm Pillar nipples and bottomed them all out before reaching proper tension. Tore the wheels down and used 12mm Sapim nipples and that did the trick. As mentioned above, you should be able to go with longer nipples to "fix" it.

14mm Pillar brass nipple on the top.
12mm Sapim brass nipple on the bottom.

2 spokes of the same length.

Bob

If you're referring to nipple breakage when spokes do not reach and engage the heads, it's anything but an urban myth. It's a well documented fact, and happens to brass as well as aluminum nipples.

There is some debate about the mechanism of failure, whether it's classic notch failure or whether corrosion might play a role, but with the exception of steel nipples, nipples can and do fail if the spokes don't engage the heads properly -- ie. 2mm or so. Nipple failure seems to be more likely with 14g plain gauge spokes vs. butted or15 spokes, so those building with 15g butted spokes are least likely to need worry about it.

To be clear, I don't say that nipples will fail, just that they may, and the risk increases with time and wheel stress. OTOH, since failure an be just about 100% prevented with proper length spokes, it's hard to justify short spoking a wheel you spend serious time and dollars building in the first place.

I don't say that nipples won't break (though it is something rare with brass nipples), but IMHO the reason why brass nipples do is because of manufacturing defects, perhaps combined with bad nipple well and nipple interfacing (and extreme stress). The main reason why I don't think too short spokes are a major cause of nipple breakage, is because the photos I have seen of broken nipples, shows broken nipples even on spokes with the correct length. So nipple breakages seem to occur anywhere on the nipple, even with correct spoke length. Of course, the usual breaking point seem to be around the flared section.

And as said before, DT Swiss 14 mm and 16 mm nipples have full thread engagement when the spoke terminates 1 or 2 mm below the top of the spoke. While the 16 mm nipple may be rare, the 14 mm is quite common, so surely they would be lots of reports with broken 14 mm nipples

Not sure if anyone have made some FE analysis to show whether alu nipples actually are strong enough for hard bicycle use (MTB, loaded touring) without breaking. I have a hard time believing, that the many broken alu nipples reported on mtbr.com are caused by too short spokes. It seems to me that alu nipples may break on even correctly build wheels, simply because alu nipples are only marginally strong enough for MTB use. So if alu nipples are inherently weak, they must surely be more susceptible to stress cracks caused by too short spokes, but that doesn't seem to be the case.

It all suggest to me, that the major cause of nipple breakage is because the nipples were inherently weak to start with because of eg. manufacturing defects. If too short spokes works as a stress riser that causes nipple breakage, one would expect to see systematic and massive failures across factory wheel batches with too short spokes, or with 14 or 16 mm DT Swiss nipples. But that doesn't seem to be the case; brass nipple breakage seems rare and alu nipples seems to break whether the spoke was too short or the correct length.

All in all I will say, that too short spokes, like 3 threads protruding underneath the brass nipple shank, is a cosmetic issue, not a systematic problem, since too short spokes are very common, but brass nipple breakage is indeed rare.

Some 40+ years ago, the US CPSC conducted serious tests on spokes and nipples as part of the formulation of the first standards relating to bicycles. Back then short spoking was SOP by many of the US manufacturers, and after the standards were in place, the bulk of short spoked wheels with brass nipples failed, while full length (2mm+ in head) passed with no issues.

It ws assumed that the major US makers would then abandon their poor practices, but they found an easier solution. They substituted long steel nipples, and these had the structural strength to pass regardless of the amount of spoke engagement, leaving them free to short spoke to their hearts' content.

Alright. Here's the deal: I measured the ERD using 2 spokes w/nipples installed,and nicks cut into each at 180mm from the bottom of the nipple head slot.
I inserted from opposing spoke holes and measured the distance between nicks using my spoke length gauge to both align the spokes and measure the distance.
This is pretty close to the method which Wheelcalg describes in their help section.
Total distance is 600mm, or 4mm more than the specs found online in several places including calculators which have the Dyads in their databases.
So, I guess I have egg on my face on this one.
However........ (read "In my defense")
I have built many,many wheels using this program, including wheels using Dyad rims. I have always used the manufacturers' published dimensions and never had this issue.
Never.
I looked at the previously built Dyad wheels in my shop, and they have NO spoke threads exposed, using the same Sapim spokes and nipples.
I reviewed the data from those builds and I used Velocity's published 596mm dimension. Had I used 600mm, the spokes would have been too long.

So, what do I do in the future **********?

Most wheels I build with spokes calculated by this program pretty much allow the spoke ends to be evened up at the base of the slots before final tensioning is applied.
That is one method I like to use to establish a baseline for tensioning. It usually proves to be pretty close as a starting point.

FWIW, I threaded a sopke through the spoke heads on the wheel that I have built with the spokes that are too short, and the spoke threads in 6 turns +/- before contacting the end of the installed spoke. The wheels were not brought to final tension, and I would estimate that there needs to be 1 turn +/- to bring then to final tension. From what I can see, that means that the ends of the spoke would not quite reach the transition between spoke barrel and head. Although it is a front wheel and I could probably "get away with it," I will order new spokes and relace. I can use the NDS spokes from the previous calculation on the DS of the rear wheel, so I only need 54 spokes + a few spares.
The program calculates 3mm longer, Is that safe, or, should I allow for the rim to shrink under tension?

BTW, the program from Professional Wheel Builder shows a schematic of the various dimensions, and the box for entering the ERD clearly shows this as the inside of the spoke hole. I guess they adjust their calculated lengths to add for nipple head height??

https://www.youtube.com/watch?v=gCXFKN9JvCQ

=8-)

Thanks, Mr,Rabbit. That information is helpful.
I am attempting to learn how to use Spokecalc in hopes that I do not encounter this problem again.