Spoke lengths - round up, or round down

[Yellow Saddle]

I also follow Roger Musson..... I guess Yellow Saddle must be wondering who the hell is this Roger Musson that keeps coming up in nearly every thread Well, Roger Musson is a guy who spent many years building mainly MTB wheels and he used to support riders in down hill competitions. He has a Engineering background and he wrote a book on how to build wheels, from the practical point of view. He does mention things like stress relieving but not in the way Jobs Brandt does. The important thing for a beginner to understand is that something like stress relieving is important in the process of wheel building and at that stage the beginner doesn't need any more information. Jobs Brandt book can be a bit overwhelming with too much information for a beginner I think. Undoubtedly it is a great book but, for me at least, it was too much to start building wheels not patient enough I guess, I just wanted to get on with it and build a wheel. The second part of JB book is the practical part and if I'm honest I think RM book is better in that aspect.

Now, going back to the point I wanted to make before I tried to explain a little bit who Roger Musson is.

I don't have RM book here with me but I'm pretty sure he would have rounded off more like this:

265.3 266 Front - down to 265 and up to 266 if no spokes found in 265
263.8 264 RDS - down to 263 if no spokes found in 264
265.4 266 RNDS - down to 265 and up to 266 if no spokes found in 265




I have followed RM advise and used RM spoke calculator for over 200 wheels and it has been spot on. I guess RM allows for the spoke stretching under tension. He does say that the lengths of his spoke calculator should be treated as a maximum and not to round up or down more than 1mm, So faced with something like 265.6 then down to 265 but a 265.7 up to 266 BUT as you know, sometimes you just cannot find the correct spoke length so 265 or 266 is fine. If I had to choose a spoke because 265 and 266 aren't available then I would go with a 264 rather than 267. I would need to alter the first stage of the building process slightly but I would be fine.

Thanks for the explanation of who Roger Muson is, but let's stick with the problem at hand.

Give me a single reason why you would want to round up, not down.

 

[Spoked Wheels]

Thanks for the explanation of who Roger Muson is, but let's stick with the problem at hand.

Give me a single reason why you would want to round up, not down.

The reason I can think of is that going too short with the spokes will make the lacing process very difficult in the last stage. The last set of spokes would be too tight already. That would mean, going back on every spoke an unwinding a turn or two.... which in turn could potentially bring another problem, nipples to loose and loosing them inside the rim.

I think you misunderstood what I said, if you read my previous post then you'll see that I do round down too

 

[Yellow Saddle]

The OP seems to be following Roger Mussons excellent book, an extract I have quoted above saying its OK to round up no more than 1mm, as this is his 1st build he doesn't have the experience to sort out any build problems shorter spokes may introduce, at 1mm or less longer they will not prove a problem, experienced builders know your spoke lengths would be fine as spokes do stretch & rims compress slightly but I always find it an easier to build if I stick close to the measurements from Roger Mussons calculator, when the op has a bit more experience he will get a feel for how much he can deviate from the norm.

I wouldn't say its bad advice to round down (slightly), but its also not bad advice to round up (slightly).

Roger Mussons calculator is designed to knock 0.3mm of the theoretical spoke length.

I have no beef with Roger and indeed, I don't know him nor have I seen his book. I want to argue from first principles, not because someone said so.

I have a problem with advice that doesn't give the OP enough information to work with. I may have been guilty too.

But lets talk about spoke length calculators first. Spoke length is calculated by using the formula for a line in space. Imagine an open shoe box. Inside the shoe box you are holding a chopstick. The chopstick is smaller than the box in all aspects. It is shorter, thinner and it's radius is less than the width, depth or length of the box. I want to know the length of the chopstick but I cannot measure it directly. There's a Chinese curse on doing so.
I'm holding it inside the box, at a skewish angle, in such a way that it doesn't touch anywhere. The Chinese Zen master has given me some dimensions to work with. He's told me what angle the stick is in relation to the sides of the box, he's also told me how high and far away from the sizes either tip is. I can now use the formula for a line in space and plug in the values and out jumps the chopstick's length.

Obviously our chopstick is the spoke and the box is the wheel. The problem is slightly more complicated when working with a wheel instead of a box. We know the hub's dimensions, we know the rim's dimensions and we know what angle the spoke will lie because we know how many spokes have to fit in the circle and at what angle (dictated by number of crossings).

Nevertheless, out jumps the spoke length. This length is the length from the point where the spoke exits the hub's hole to the end of the inside of the nipple's slot. I call it the mathematical, or nominal length. We have to adjust this length because of several reasons:

1) The spoke stretches under tension. A thicker spoke stretches less than a thinner spoke.
2) The rim compresses under tension. Big rims compress more than skinny box section rims.
3) The nominal length almost always includes a fraction and spoke sizes aren't available in fractions.

We have two constraints in our flexibility to change the length.
1) Ideally the spoke should not protrude past the slot in the nipple. This is so that it can be removed with a screwdriver if the flats stripped. A secondary reason is that if you use a speed driver, a too-long spoke ejects the driver before you've wound up much tread and it slows the job down.
2) The variance we have in how far in or out the spoke can be screwed in is small. At the upper end is in point 1 above. At the lower end it is thread showing below the nipple once the wheel is tensioned.

Now, if you were to round up, you would potentially end up with a spoke protruding. That's not the end of the world, but rounding down would be better for the reason stated above, as well as professional pride.

The OP asked whether he should round up or round down. I said ALWAYS DOWN and then indicated how and why. Some of you said it depends. Well, depends on what? If it depends on the spoke calculator, then we should only answer the question once we know what calculator was used. Even if the calculator is cited, we still don't really know how its engine works and how it compensates, if indeed it does. I think a good wheelbuilder will want total control over the calculation and work with a raw number rather than a mystic number. This way he understands what his rounding will do to the end position of the spoke end.

Further, the conditions I've cited are based on my calculations for the strain (elongation) of 1.8mm and 1.5mm double-butted spokes under 1000N tension. It is a formula with no if and buts and maybes. This really isn't black art, it is science.

 

[Yellow Saddle]

The reason I can think of is that going too short with the spokes will make the lacing process very difficult in the last stage. The last set of spokes would be too tight already. That would mean, going back on every spoke an unwinding a turn or two.... which in turn could potentially bring another problem, nipples to loose and loosing them inside the rim.

I think you misunderstood what I said, if you read my previous post then you'll see that I do round down too

If you round down by the number I've quoted, you won't find problems with the last set of spokes at all. I manage to wind up all the nipples right up to the end of the spoke's thread (on the hub side), using just a speed driver. As you know, a speed driver exerts almost no torque.

 

[Spoked Wheels]

If you round down by the number I've quoted, you won't find problems with the last set of spokes at all. I manage to wind up all the nipples right up to the end of the spoke's thread (on the hub side), using just a speed driver. As you know, a speed driver exerts almost no torque.

I don't think I read the post you are referring to
so I will look for it. But, I already said that I would go down to nearly 2mm with no issues so I'm guessing here that you are getting exited sbout maybe 3 or 4 mm below the number given by a spoke calculator. Can I ask you to give me a reason as to why I shouldn't keep my spokes in the range I already mentioned? I remind you that it appears that I'm not the only one using spokes in such range but more importantly, my spokes at 120 - 130 kgf end at a place that I'm happy too see, they are not too long that will pose a threat to tubes.
So, if my spokes are fine and you are saying I could go 3 or perhaps 4mm shorter. What is the advantage?

BTW, I don't use a speed driver, just a nipple driver.

Edit:

After reading Yellow Saddle post. I think I read that before but a few days ago.

Not as much difference to the amount I'd be prepared to round down as I first thought. I would not pick spokes that short automatically though, only if my preferred length were not available.
I don't see the problem with rounding up 0.3 of a mm max. I think you are stressing yourself out too much over 0.3 of a millimeter

To summarise my views:
I would automatically go up up to 0.3 mm or down 0.7mm - having said, I would go a further 1 full millimetre down if my preferred spoke length is not available.

 

[Venod]

The OP asked whether he should round up or round down. I said ALWAYS DOWN and then indicated how and why. Some of you said it depends. Well, depends on what? If it depends on the spoke calculator, then we should only answer the question once we know what calculator was used. Even if the calculator is cited, we still don't really know how its engine works and how it compensates, if indeed it does. I think a good wheelbuilder will want total control over the calculation and work with a raw number rather than a mystic number. This way he understands what his rounding will do to the end position of the spoke end.

I suggested the spokes assuming that the OP had used Roger Mussons calculator (he may not have but as he quotes the book its most likely what he used) this removes 0.3mm from the theoretical length, the formula is fully explained in the book.

based on the ops calculated lengths (assuming Mussons calculator was used) we get the theoretical lengths of

292.7 + 0.3 = 293
291.3 + 0.3 = 291.6
293.6 + 0.3 = 293.9

Very close to what I recommended and in my opinion a lot easier to work with than the shorter options you suggested which would give spokes almost 2mm shorter than calculated.

 

[Smurfy]

Well, then, explain why it is poor advice and tell me when it is OK to round up.

Telling me that so-and-so says so is not a good enough argument. You have to find a flaw in my reasoning. The actual spoke calculator is irrelevant in the matter, it is an understanding of how much a spoke stretches at 1000N of tension, that dictates the starting length and last time I looked, not a single spoke went shorter under tension.

Whether you should round up or down depends on:
1. if you have access to spokes in 1mm or 2mm increments
2. which spoke calculator you used
3. how the spokes and nipples are threaded

 

[PpPete]

1) Ideally the spoke should not protrude past the slot in the nipple. This is so that it can be removed with a screwdriver if the flats stripped. A secondary reason is that if you use a speed driver, a too-long spoke ejects the driver before you've wound up much tread and it slows the job down.

As ever, some class nuggets of utter BS in amongst mostly good sense from Yellow Saddle:

"Ideally the spoke should not protrude past the slot in the nipple" - can't argue with that ... but we don't live in a perfect world.
"This is so that it can be removed with a screwdriver if the flats stripped". Oh come on ! If it's frozen on the nipple so hard that you strip the flats you ain't ever going to move it with a frikkin screwdriver.
"a too-long spoke ejects the driver before you've wound up much tread and it slows the job down" Big deal ! we are (mostly) amateur wheelbuilders on here and the extra 5 minutes to give an extra turn or two around the wheel with the spoke key is not relevant.

I don't think which spoke calculator you use makes a whole heap of difference. RM introduces a 'fudge' of 0.3mm - again big deal - it really is not that easy to measure the ERD to that accuracy, and taking the manufacturer's word for it is (potentially) even worse.

What I would suggest is that if, on a dished wheel, the theoretical results came out at, say:
290.1 drive side
291.9 non-drive side
and you had spokes only at 2 mm (even number) increments you would make life a lot easier for yourself in the build process if you chose
290 drive side
292 non-drive side
Unlike with 'always rounding down' the difference will get you close to good dishing as soon as you've 'taken up the slack'.....which if you've rounded one side down a lot more than the other you'll quickly find is not the case.

 

[Tim Hall]

<snip>

1) The spoke stretches under tension. A thicker spoke stretches more than a thinner spoke.

This seems counter intuitive,or as we used to say, wrong. Surely, the thinner will stretch more (until it breaks) for the same amount of tension?

 

[Yellow Saddle]

This seems counter intuitive,or as we used to say, wrong. Surely, the thinner will stretch more (until it breaks) for the same amount of tension?

Sorry, mistake. It is obviously the other way round.

 

[Yellow Saddle]

As ever, some class nuggets of utter BS in amongst mostly good sense from Yellow Saddle:

Why can't you just be civil and debate or disagree?

 

[PpPete]

Why can't you just be civil and debate or disagree?

Touched a nerve there did I ?
I'm sorry but I find the pomposity, arrogance, and condescending tone of many of your posts frankly unpleasant.
I wouldn't normally mention that. Your stylistic choices are your own and if you want to come across as a pompous ass that's up to you and I'll let you get on with it.

That said - I agree technically with much of what you write.
You'll just have to put up with the fact that my dislike of your 'tone' prompts me to pick holes in your arguments.

Now I've answered your question...- yes please let's debate the technical points I made.

 

[Smurfy]

As I understand it, there's quite a few spoke calculators around, and some of them give different results. I prefer spocalc express, specifically because it gives me unmolested values. That way I can consider all the various factors, and then decide whether to round up or down. The Sapim online calculator is one to watch, as their definition of ERD is reportedly slightly different from others. Of course all of these calculators give the exact right answer, providing that you understand what they are asking for, and what assumptions/corrections, if any, have been applied to get the answer (e.g. flange hole diameter, flange thickness, automatic subtraction of 0.3mm).

It's also worth testing the threading of the spokes and nipples before using. Some combinations of spoke and nipple will permit you to have 2mm of spoke thread protruding above the nipple, while others will bottom out the threads long before that point. This can affect the decision of whether to round up or down, particularly if you only have spokes available in length increments of 2mm.

 

[Yellow Saddle]

As I understand it, there's quite a few spoke calculators around, and some of them give different results. I prefer spocalc express, specifically because it gives me unmolested values. That way I can consider all the various factors, and then decide whether to round up or down. The Sapim online calculator is one to watch, as their definition of ERD is reportedly slightly different from others. Of course all of these calculators give the exact right answer, providing that you understand what they are asking for, and what assumptions/corrections, if any, have been applied to get the answer (e.g. flange hole diameter, flange thickness, automatic subtraction of 0.3mm).

It's also worth testing the threading of the spokes and nipples before using. Some combinations of spoke and nipple will permit you to have 2mm of spoke thread protruding above the nipple, while others will bottom out the threads long before that point. This can affect the decision of whether to round up or down, particularly if you only have spokes available in length increments of 2mm.

I think you make a good point and particularly the sentence I highlighted. It would help any builder to understand exactly what the calculator is doing or not doing.

As for the various combinations of spokes and nipples, all the major spoke brands use a 9mm long thread. This includes SAPIM, DT, Wheelsmith and Pillar. Have you come across any with different thread length. In addition, nipples with exactly the same thread characteristics (for a given length nipple) include SAPIM, DT, Pillar and Lilly. Which ones have you come across that bottom out with different results? Perhaps I should ask about the particular combinations.?

 

[rb58]

Thanks everyone for all the advice here. Plenty to think about. I shall report back on how it all goes.