Kiroskka

This had been something that I had been meaning to do for a while now, but never got around to until recently. Riding self built wheels just has such a great appeal
to me. I wanted to build a wheelset that would add a better classic touch than my other wheels. The wheels I had been using always looked a bit too out of place to me.

The rims were the first thing I decided I would buy. After some pondering and then some looking around, I accquierd some NOS Mavic MA-2 rims.




Upon initally inspecting them, I could not help but laugh when I saw the original price tag still stuck on.



For hubs, I decided to go with the current Dura Ace hubs from Shimano.




To obtain the proper spoke length, I made sure to measure everything myself. I used the method described on Sheldon's site to find the ERD. I measured the ERD in three places
on each rim, coming out to 613mm on both rims. I was pleased that both rims were round and flat to start out.

I compared my measurements with the measurments Shimano has listed. Shimano's numbers still just don't make any sense; I don't know what the hell they were measuring.
https://www.shimano.com/publish/conte...ub%20Specs.pdf

My measurments, using a vernier caliper, came out as the following:

HB-7900 (front hub)
flange diameter - 38mm
hub center to flange center - 35mm

FH-7900 (rear hub)
flange diameter(L) - 44mm
flange diameter(R) - 45mm
hub center to flange center(L) - 34mm
hub center to flange center(R) - 18mm


Gerd Schraner's book "The Art of Wheelbuilding" was very helpful. I used his spoking method for my build. The illustrations were very clear, concise, and understandable.I also
used Sheldon's wheelbuilding page as a secondary source for everything else once I had finished all the spoking.

I went ahead and made sure the orientation of all the labels was "correct"; it seems like one either rolls his eyes at such a matter or holds it to some importance. I like
the aesthetics of it. Even if it isn't noticable when riding, I still like the idea of it. Besides, it was a very simple task.

I called it done when I got the true within 0.3mm; it became a bit too much trouble for me to get any lower. I had some dial indicators, but they were not as useful as I
thought they would be, though very cool to look at. The rear wheel took twice as long to complete. I lost track of how long it took me to complete both wheels, but it was definately
over 7 hours.

I used DT competition spokes throughout.

300mm front
299mm NDS
298mm DS

32 spoke count
3x cross
symmetrical



I am 135 pounds soaking wet. So, yeah, I'm expecting a lot of milage out of these.

I think I'll stick to personally building all my wheels from now on instead of sourcing the work to others. It's a lot more rewarding.

skydog6653

Nice job! How much $ was the tool outlay? How many wheelsets will it take to "break even". I know this was a labor of love and not one of economics, but I'm curious.

Retro Grouch 

Oh dear! Now you've got the disease too. There's no known cure for it so you might as well just go with it. Mrs. Grouch refers to wheelbuilding as "knitting for men".

Nice job by the way.

BentLink

Great components and a little attention to detail got you off to a nice set on you first try! Enjoy.

Kiroskka

I think I spent a littel under $300 for all of my tools. One could build a good set of wheels for a lot less; it was all about making the job easier. The labor cost for a wheelbuild at my LBS is $45

marcusbandito

Nicely done. Have fun with the new skillz

blamp28

Retro,

I literally laughed out loud at that one! Oh and I resemble that remark. I have often commented that wheel-building is therapeutic for me. or that I'm overdue to build a set of wheels but I have no current need. That's just funny and true.

Kiroskka,
Nice job! Enjoy your new obsession. I hope it brings a balance to the tensions of life.

mrrabbit 

Not surprised at the Shimano weirdness....

They measure not to the center-of-flange for offsets - but somewhere past - just like Campagnolo. Campagnolo is pretty easy to adjust - just subtract 1.6mm. Shimano seems inconsistent where on the outside they are measuring so you can't just subtract a regular number...

For Shimano, I've have to include their spec in my spreadsheet - then update as I encounter them...covered about 2/3's, still have a way to go...

https://www.mrrabbit.net/wheelsbyflemingapplications.php

Also, myself and a few others here have caught a few flange diameter mistakes on their PDF - one affecting a disc one and another affecting some Dura Ace hubs.

Question for you:

Are you sure that was FH-7900 rear? (Not FH-7850?)

=8-)

FBinNY 

The center to flange measurement is the least critical of the measurements.

ERD measurements affect spoke length on a 2:1 (diameter to radius) ratio, ie. +2mm ERD means +1mm spoke length

Flange diameter affect length -2:1 for a radial pattern, (-2mm dia. means +1mm spoke length), and not at all with a true tangent, ie. 4x 36h, or 3x 28h, and to varying degrees for other patterns.

Center to flange affects length on a 10:1 ratio. a 1mm difference changes the spoke length by 0.1mm

The exact ratios vary slightly within the general bands I described, but the concept holds. I offer this to spare folks the effort of agonizing over details that don't mean anything in the real world. For example a 2mm difference in CTF changes the spoke length by 0.2mm which doesn't factor when you're going to round to 1mm increments anyway.

Kiroskka

Yes, definatley 7900

I think I'll just pretend that hub measurments straight from shimano don't exist from now on. ha

It's nice to know I can be less OCD about some measurements, but I don't think I can help myself.

mrrabbit 

When a spoke length calculator is used - it is averaging the center-to-flange distance for purposes of spoke length calculation. The reality is that you are really calculating 2 center-to-flange distance for EACH side of the hub.

For example - if the center-to-center_of_flange distance is 34.00mm for a modern day hub - what you really have is roughly 36.6mm for outside spokes and 31.4mm for inside spokes - or something like that.

Futhermore - that slowly starts to change the further you move the flange out assuming the flange diameter stays the same. The 10:1 ratio you introduced actually drops to 9:1...slowly but gradually...when working with CTFC's of 38mm and 40mm like with the old Maillard's...

What happens when you are that far out? You can get as much as a 1.00mm difference in spoke length between the outside and inside spokes for the same flange. It can cause a situation where half of the spokes can run short or run out of threads for a given nipple.

Keyword is can...pretty rare though...we're talking extremes here...

My problem with Shimano is they are measuring to the outside - making the insides spokes longer than necessary - and making the outside spoke even more longer than necessary. I want the average - which is what I get when the measurement is to the center of the flange.

Overall it's really not a big deal...unless you go all elbows in for radial...or when you have a flange that's way way out there....45.00mm is the worst I've seen - electric motor hub I think...

American Classic and Bullseye have had CTFC's of 40.00mm to 42.00mm in the past...

Basically it'd be nice if Campagnolo and Shimano would do the same as most of us when we are measuring hubs for spoke length calculation purpose - reference the center of the flange.

I can dream can't I?

Now back to ERD...

=8-P

mrrabbit 

Kiroskka:

Normally I input what I measure - but if okay with you I'm going to go ahead and update my spreadsheet that I publish noting "kiroskka" of BikeForums.net as the source for the FH-7900.

=8-)

FBinNY 

Sorry Mr. Rabbit, Trix are for.... (sorry, couldn't resist)

With the understanding that we're splitting hairs, and IMO it doesn't matter in the scheme of things, I disagree with the above.

There isn't any difference in spoke length to the inside or outside of the flanges since the spokes don't take a direct route to either. The spokes are (usually) laced over/under at the last cross, which effectively averages the two distances. If building mirrored, ie. pulling spokes to the outside of both flanges, any long spoke/short spoke pattern seen at the rim is the result of hub twist, (shell is rotated slightly with the respect to the rim) and nothing else.


Shimano and Campagnolo are correct in their analysis, though it only matters to the truly anal. As i said before, neither spoke runs directly to the flange. The actual route passing it's crossed partner and back to the hub is longer than the direct route, so these folks use a CTF dimension which compensates for where the same flange would be if the spoke ran a straight line, passing the cross point continuing unbent to the flange. You can see the effect yourself if you draw exaggerated cross section of the spoke's route.

In any case, all these small variations, are one order less than the 1mm increments spokes come in and therefore immaterial. Every spoke calculation has an adjustment for these small constants, along with that for the desired final height of the spoke in the nipple.


--------

Computer spoke calculation formulas are all similar, starting from half the rim diameter, using simple trigonometry to then compensate for the effects of flange diameter, cross, and center/flange distance.

But length calculating calculation isn't new no a dark complex science. I still prefer to use my old standby that I used long before hand held scientific calculators and on line computing made the current formulas practical.

Start with 1/2 NSD (my term for diameter at the nipple seat), subtract 1/2 the distance between holes for spokes going in opposite directions from the flange (equals the distance short of the center of the hub), add 1mm for every 10mm CTF, add for desired height in nipple. This gives me an accurate, dependable spoke length calculation without using complex math or electricity. The built wheels consistently end with the spokes ending exactly where i expect them to.

mrrabbit 

"I disagree with the above."

You sure? Draw a picture...or ping Jobst Brandt, Mr. Sutherland, Damon Rinard, etc...

When you measure an offset to the center of a flange - that number you come up with is an average of the outside and inside spoke offsets assuming the typical half-inside/half-outside lacing.

You can do it yourself by grabbing a hub - measuring the offset to the center of the flange - and then sticking in two spokes and then measuring the offset to the center of of each spoke just before the elbow bend begins. You'll end up with somewhere in the ballpark of 5.2mm to 6.0mm from one spoke to the other as well....

I can provide a picture if you'd like...

What I've stated is nothing new...nor anything I've invented myself...folks much older than I included those noted above are familiar with this - and knew long before I started building wheels...it's old news.

=8-)

mrrabbit 

Also, open up my spreadsheet.ll

It has the hub advantage of displaying ALL hubs in the database all at once in one long continuous page for "spokechart". Pop-in a theoretical 16 and 3 at the top - and a 612 for diameters.

Find a hub class/grouping where for the same flange diameters, you can see slight changes from 17 to 18 to 19 to 21.50...etc. I think you'll see your 10:1 ratio.

Then find a hub class/grouping where for the same flange diameters, you can see slight changes from 36 to 37 to 38 to 40 to 42...etc. I think you'll see roughly a 9:1 ratio.

Then do another experiement...for 38s, 40s, 42s, etc....find related offsets that are 2.6 - 3.0mm higher and 2.6 - 3.0mm lower. You'll see on average a 1.0mm difference in spoke length.

Then picture what a difference it makes if you suddenly decide to go all-elbows in or all-elbows out - especially if the spoke length offset reference was to the outside - and the wheel you are building is going to be elbows in...

Not too bad for 10mm and 12mm nipples - but can be a party spoiler for 14mm and 16mm nipples with restrictive threading.

=8-)

FBinNY 

yes, I'd like.

mrrabbit 

I took several photos and added a graphic.

1. Blurry - but measured to center-of-flange.
2. One showing measurement to inside-of-flange.
3. One shoring measurement to outside-of-flange.
4. Graphic summarizes the measurements in photos...
5. Added photo of Maillard 700s with extreme flange layout - just for eye candy.

Reason we measure to center-of-flange is so that we can provide a little extra length for the outside spokes without adding too much to the inside spoke length. 99.99 percent of the time - this works out just fine for traditionally laced wheels.

The center-of-flange measurement just so happens to be an average of the outside and inside spoke offsets...and the distance between the two typically is around 6.0mm for your typical alloy hubs with 3.2mm flange thickness.

We are describing 3 triangles here - each with different angles - which means they are not proportional triangles. One is a rough average of the other two.

We can use the ratio of 10:1 you mentioned earlier as a rough indicator of what happens to spoke length when you move a flange 1.0mm....but once you get past 40.00mm of offset...you'll see some creep to .11mm, .12mm, .13mm...for a given flange size.

Use the spreadsheet...play around...focus on American Classic (old), Bullseye, and Maillard. Then again, I might be wrong - it could be the flange diameter creating the creep.

Check for yourself...

When I build electric motor hub wheels where the customer says, "all elbows in please...", I'll chop off another .25mm of spoke length of the calculated result knowing that I no longer have to leave it in to accommodate outside spokes since they won't be present.

I do it just because I can...spoke machines give you that liberty.

(Electric motor flange thicknesses are often 3.4mm to 3.6mm...from what I've seen so far...)

=8-)

FBinNY 

I see the picture, but it doesn't make your point for typical builds. What you miss is that the spokes don't take a straight path to the rim. They meet and cross over each other inside over outside enroute. The difference in spoke length would only become an issue for radial wheels.

If you sketch an actual pair of spokes running from flange to rim, you'll see that the distance inside to outside equalizes somewhat if not completely. You'll also see that both spokes are longer than if they were straight (shortest line between two points). (Shimano/Campagnolo fudge factor)

In any case (for all non-radial wheels), when the spokes are tightened to the same height, any difference between pulling and forward spoke length is equalized by the hub turning a few degrees. If the flanges are mirrored, this would equalize any difference in pulling v. forward spokes.

This equalization by hub rotation is a trick we capitalized on in emergencies if the correct spokes weren't available. For example if I needed 296mm and didn't have I would build with 298 and 294, and both would end up the same height in the rim, as if by magic.

The problem here is you're looking at a bunch or trees, and I'm managing a forest.

mrrabbit 

It seems like you are disagreeing - just to be able to disagree....

1. I never said the spokes don't take a straight path to the rim...I'm talking about spoke length calculation with a focus on offsets and the distance - and how it affects spokes and wheel builds...which is very little.

2. You talk about equalization...I'm using the word average...however inside and outside spokes are covering a different distance, a difference that varies with offset and flange diameter.

3. The calculator provides one spoke length per side when you provide one offset per side. However, the distance being covered by the spoke inside and outside is different. This is why oftentimes drive side spokes on a wheel with a very small drive side offset will arrive at the same point in a nipple, whereas non-drive spokes covering a large offset will vary every other spoke. The angle is so small on the drive side that whether interlaced or not - inside and outside spokes will arrive almost at the same point on average. However the angle on the non-drive will be such that the distance covered by inside and outside spokes becomes slightly more noticeable due to the observable inconsistent landing points of spokes ends in the nipples with a pattern of sorts.

This effect becomes most noticeable in rear multi-speed hubs with non-drive offsets in the high 30's and low 40's. (Maillard Helicomatic 126mm rears are a classic example...)

4. "The difference in spoke length would only become an issue for radial wheels." We are practically saying the same thing...why the "disagree"?

We are not looking at trees or a forest...we are looking at offsets, angles, triangles and distance. When two angles change (at hub and rim) - the distance changes. For practical everyday purposes though - when we measure to the center-of-flange - we are creating one triangle that represents the rough average of the other two triangles formed by outside and inside spokes. It keeps things simple, practical, and as I stated earlier, 99.99 percent of the time - it works.

I just don't like Shimano's and Campagnolo's bias to outside measurement cause it makes for a spoke length that's a tad longer than necessary - plus once again we get multiple parties measuring the same thing but with different references. I know Campagnolo is simply a reference to the outside of the flange thickness (1.6mm), but Shimano? I can't be sure, because no matter how hard I try I can't get a consistent number for the difference between on center and their number...

So I have to wait until the hub is in my hand...or someone else provides convincing numbers...as kirosska did.

All Shimano has to do is provide a graphic indicating where they aim for offsets and I'd be happy.

=8-)

FBinNY 

Not quite, go back to my original post (#9) about the relative significance of the three measurements, which you felt needed correcting. So maybe the pot is calling the kettle....

I believe that you're exaggerating very small differences making the immaterial seem material, but we'll leave it there. I think I've prefaced my posts on this as splitting hairs, or not material in the scheme of things.

Here we agree that the difference is very little.

Here we disagree. I believe that the interlacing causes spokes to take routes that are essentially equal in distance from rim to their respective side of the hub. With equal tension, each spoke has the same deflection as it's crossed partner, and the point of cross will be such as to mitigate most or all of what would have been a difference in length.

You keep saying this, and that's the crux of our disagreement.

I understand you don't like their analyses, and I don't either since I consider anything but actual raw data a complication in published specs. (remember I'm the Joe Friday who doesn't like nipple head considerations in ERD specs.) I was simply explaining that their method was a (unnecessary) way to compensate for the longer spoke route.

100% in agreement. Since everybody insists in "correcting" their raw data to compensate for various factors without saying whether it's raw or adjusted, I find all published date useless. You'll note that I still calculate spokes based on an old algorithm that doesn't include flange diameter, or the number of crosses, but a direct measurement of the locations of both ends of the spoke, corrected, within reason for flange offset. Things that cause minor differences of an order of less that what I'm rounding to anyway are disregarded as immaterial. It's stood me well for 40+ years, and makes me immune to errors caused by nice people "helping" by correcting their data.

BTW- my last post on this, I think we've both exhausted the subject and readers can read and draw their own conclusions as they prefer.

davidad

Girls, girls you are getting each other's feet wet.

davidad

The wheelset looks good. I like the pulling spokes with the head in. I read that Mavic built up a number of wheels with the different variations of spoking and found that the way yours is set up was the most durable.

FBinNY 

Oh No!!! don't tell disciples of Mr. Brandt, we'll be here forever.

BTW- I've been building the same way for 45 years (pulling spokes outside (head in), but don't feel that the opposite is wrong, just different.

mrrabbit 

1. Why are you claiming I'm exaggerating when from the get go WE BOTH AGREE the impact is minimal...

It's like criticizing a seller who posts a beat up Masi on Craigslist with a low price noting the poor condition - for posting a Masi in poor condition.


2. Distance is distance. A distance penalty equally applied to both is going to affect both the same. Both now have to travel farther.


3. Math is math:

[Begin Math Example By Using Inside Spoke Offset and Outside Spoke Offset for the Non-Drive Side of a Maillard 700 Series Rear 126mm]

C:\SPOKEC~1>spoke
Enter the number of spokes on one side:
16
Enter the number of crossings to be used:
3
Enter the rim diameter in millimeters:
612
Enter the hub diameter in millimeters:
44
Enter the offset in millimeters:
37
Your spoke length for this side of the wheel is: 299.560394
C:\SPOKEC~1>spoke
Enter the number of spokes on one side:
16
Enter the number of crossings to be used:
3
Enter the rim diameter in millimeters:
612
Enter the hub diameter in millimeters:
44
Enter the offset in millimeters:
43
Your spoke length for this side of the wheel is: 300.357849
C:\SPOKEC~1>

[End Math Example]

Notice the difference of approximately .74mm in length for the inside and outside spoke.

Impact? For typical builds with 10mm nipples back in the day and 12mm nipples today - not much.

Interlacing doesn't improve one or the other - they'll BOTH just have to "crawl" and extra micron or two. Once again, not a big deal.


I provided raw numbers and graphics FBinNY...and this isn't the first time. Sometimes I think your're in a rush to disagreement...cause sometimes I find myself thinking, "Disagree? Doesn't sound like it..."

=8-)

Welp that's all for now...

=8-)

FBinNY 

Because you are. If you agree that the difference is minimal, why the F*** did you feel you needed to jump in and correct what was a generalization (and presented as such) of the analysis with a bunch of picayune details. Does it matter if the change ratio for spoke length to CTF distance is sometimes more like 9:1 than 10:1? Will it in anyway affect the final spoke length calculation?

When someone says the sky is blue, I don't pull out a Pantone color chart and say it's not exactly blue, some is this shade and some is that shade, and some is more gray than actually blue.

Like so many of the debates on this forum, the most words are wasted arguing about fine points that are immaterial in the scheme of things. My post #9 was simply to lend a bit of perspective so that those measuring could spare themselves a bit of unnecessary grief, and focus on what matters. You, however, insist on focusing on small differences that will disappear later on anyway, and therefore, as we both agree don't matter.