Hi everyone,
I’d be looking for some wheel building advice. I have recently built my first pair of wheels and I’m having some issues with the spoke tension dropping after a few rides.
Here’s a few pieces to the puzzle:
I have laced them radially in the front and 2 cross in the back
For spokes I have used Sapim CX rays. I have used Roger Musson’s spoke calculator and had the people at Light Bicycle calculate the spoke lengths independently. Both came to the same result
I am using Sapim polyax brass nipples
For spoke prep I went old school with boiled linseed oil and left the wheels to dry for about a week.
I have bought the park tool TM-1 tension meter
I have tensioned them per the manufacturer’s recommendations ~110 kg in the front and ~130 on the drive side in the rear.
Spoke tension ratio for the rear is 100%:52% (drive side to non-drive side tension)
During the build I have stress relieved the wheels multiple times by squeezing the spokes and carefully stepping on the spoke crossings
I am around 73 kg and a good chunk of my riding is commuting with a backpack of maybe 5 kg. The commute includes some light gravel
I am running my tires with inner tubes
Considering all of the above how could I fix the issue of the rear wheel not retaining the tension (the front wheel is mostly fine)?
Since I cannot change the number of spokes and assuming that my wheel building technique is sound, the only potential fix to the issue that I can think of is using thicker spokes.
Would that help? If so which spokes should I go for and where? I have read that some people use the combination of Sapim CX Sprint for the rear drive side and CX Rays for the non-drive side. I could also go for complete CX Rays for the rear or something even burlier like the Sapim Race.
Most likely the spokes are losing tension because the non-drive side tension is too low. The hub geometry is the main culprit and there’s not much you can do about that. You could try a rear rim with an offset (and new spoke lengths) so there is better tension balance and therefore a higher tension NDS but with the rims and hubs you have, the only option is to crank up the tension to what the rim bed can bear. It’s the rim bed strength and the hub flange strength that govern the max tension, not the spokes themselves (unless you’re anywhere near the yield point, which is somewhat unlikely).
You don’t say if you are trying to run tubeless tyres or not but if you are that will be making things worse as tubeless relieves spoke tension more than tubed. Trying to stop things moving with Loctite will not work in the long run, I’m afraid.
The rear non-drive side which is the worst is currently at around maximum 50 kg, some spokes considerably less. The park tool conversion table does not go that low.
Since the build up with the linseed oil, this would be the first re-truing. Before the linseed oil, I used some medium viscosity engine oil, as per recommendation of the Roger Musson wheel building guide and they never really held the tension. So there I’d have a loss of maybe 10kg after a couple of rides and that would be recurring without really settling.
Now with the linseed oil it’s been a lot better. I’ve built them up about a month ago and put around 460km on the wheels. This is the first time now, that I have to re-tension them. What certainly didn’t help is that I did a long ride where I got a flat tire with around 100km to go. Unfortunately after fixing the flat, I could not get the tire to sit properly with just my little bike pump and so I had a pretty bumpy ride for the last 100km. I’m sure that didn’t help with the spoke tension.
So now I’m wondering if the spoke tension loss was just related to that and would not happen in normal conditions, even with a bit of gravel or if the setup with my weight, the number of spokes and the CX Rays is underdimensioned anyways and it’s going to loose tension sooner or later.
Good point about the tires, I am not running tubeless.
I’ve read that a low tension non-drive side is a likely culprit and already gone as high as the rim allows, so I’m afraid that this is not an option.
You’re of course correct, a thicker spoke would not equal more tension but I would think that it would still make for a stronger wheel. From what I understand a wheel looses tension by the spokes vibrating. I think how much a spoke vibrates is both a function of the tension but also the cross section.
So while a thicker spoke would not make for more tension, I still think it would make for a stronger wheel.
You say you used brass nipples, but what kind? I’ve had great luck with DT Swiss ProLock nipples, which include an adhesive in the threads. They take a bit more effort during the initial build, and you need a spoke-grabber to prevent windup, but once set I’ve had zero issues with loosening. DT Squorx ProLock Brass are my go-to, the Squorx head makes the build really easy.
Ah yeah, I forgot to state, I’m using the standard Sapim Polyax nipples (updated the original post). So that’s without any kind of extra retaining technology be that deformed threads or thread locker. I have a spoke-grabber tool from Sapim so I could try and switch to the DT Swiss ProLock. Good to know that you’ve had good experiences there. You’d then use those without any other oil or spoke prep, correct?
Hmmm, no, not vibration. Rather, when the wheel is loaded e.g. by a bump the hub moves down towards the bottom of the rim, reducing the tension in those lower spokes to zero. Thicker spokes reduce how far the hub moves because greater csa means less elongation of the upper spokes.
Would it be a ‘stronger’ wheel? Laterally stiffer, yes. Will it fix the problem? Maybe.
I’m no pro but have built a dozen or so wheelsets.
Perhaps this “considerably less”, together with a low 50kg is the problem here?
If you have Roger Musson’s book (if you don’t, I highly recommend it!), look for the part where he talks about equilizing spoke tension.
It’s uneven spook tension that makes (the lowest tension) spokes loose tension while riding.
What variation are you getting (between spokes on the same side of a wheel)? I think he says it should be below 15%, and with a little practice you can get <10% and they never loose tension.
Also, and this is a long shot: are you sure you’re not getting spoke wind-up? This is where the spoke twist under tensioning, easily done with bladed spokes. When they unwind while riding they loose some tension.
2nd also; be sure to check DT Swiss’ advice re: radial spoke patterns. Some hubs can’t handle the stresses of radial spokes.
This is good advice. I try to get all similar spokes around the same reading on the Park tool and with C-xray spokes no lower than 3 - 4 on the nds. You can sacrifice some offset when dishing the wheel to improve nds tension too if you’re desperate, having a wheel that’s a millimetre or too closer to the nds side is better than a wheel that falls apart. Just re-centre your brakes.
i’m sorry to be the guy but noone pointed this out so far - the park tool tensiometer is not very precise nor accurate tool to the point I would advise not to use it. Have a LBS with a proper tool take a look.
What about the hub? Is it new old stock or is it being reused? I’m not sure by how much and if it would make a difference but aluminium loses its properties as it ages. 17 years is a lot. If the hub is being relaced, the spoke heads could be just settling deeper into the previously stretched holes.
I also built handful of wheels with similar parts (not that old though), I’m seeing a similar nds/ds tension ratio when building dt swiss hubs with round rims (princeton carbonworks style serated rims have been better) but I’m not experiencing repetitive tension loss. I use sapim secure lock double square brass nipples, no spoke prep and mineral oil for the nipple head/rim interface
I did buy and study the book by Roger Musson and I did equalize the spoke tensions. I verified that they are all about equal using both the tension meter as well as the sound they make when plucked with a guitar pick. Both methods also aligned relatively, i.e. the highest pitch had the highest reading on the tension meter.
Regarding the variation, I got it to a point where they all read between 6 and 7 on the tension meter. So I could have brought them closer but between 6 and 7 would be at least within ± 10%.
I don’t think I got any spoke wind up. Since I’m using bladed spokes this should be easy to spot. I got the bladed spoke holder by Sapim and always made sure to grab the spoke when turning the nipple.
Good point about checking if the radial pattern is supported. I only run that in the front and the hub is specifically designed for this. The original DT Swiss sticker says “DT Swiss 240s radial”. Previously, the hubs were laced to a pair of Zipp tubular wheels. Since I did not want to deal with tubulars, I went for the rebuild in the first place (and of course because I wanted to learn how to build wheels).
I kept the original lacing pattern for both wheels, so radial for the front and 2 cross on both sides in the back. I also followed the advice by Roger Musson and kept the spoke orientation (pulling vs. pushing, heads-in vs heads-out) the same so that the in the rear the new spokes follow the indentations that the spokes on the previous hub made.
@Nick_Coward when I built the wheels I got all the spokes to around 6-7 on the park tools meter so that should be alright. Good point though about sacrificing a little bit of the dish for some more nds tension.
Ah, right, of course. Thanks for the explanation. It now clicked for me for the first time. Of course a thicker spoke means a stiffer spring in any kind of direction. However, 100 kg (of however much) of tension is 100 kg of tension regardless of the spoke. So no matter how thin of thick the spokes, if you apply the same force in the opposite direction of their tension they go slack under the same load. Totally makes sense
I can see now how laterally, this would make for a stiffer wheel but not vertically.
Interesting to hear that the park tool tension meter is not precise enough. Out of curiosity, what do you use to build wheel? Do you go by pitch and/or feel alone and then have a LBS check? Luckily I have a great LBS so I’ll have them have a look.
In my experience with the park tools meter, it seems accurate at least in relative terms. When I’m equalizing spoke tensions instead of measuring each spoke each time, I pluck them with a guitar pick. I have found that this gives me a very nice and repeatable pitch, much better than using my finger (nail). What I’ve found is that the pitch is consistent with the park tool readings, i.e. the highest pitch gets the highest rating. It could of course be though, that the tension meter is off in absolute terms so I’ll have my LBS check next time for sure.
The hubs are being reused. As per the Roger Musson book I have made sure to run the spokes for this build along the indentations already created by the previous one. It could of course be though that there’s some more settling of the spoke heads.
Interesting to know that you’ve had good experiences using the sapim secure lock double square nipples. I’d be curious, how hard are they to turn? Do you need the key for the square interface on the back of the nipple or can it be turned just fine using the normal nipple interface? Is there a key for the square interface that you’d recommend?
Where I work we use Hozan C-737 tension meter in combination with a tension calibration jig which is basically a crane scale in a metal frame with attachment points for spokes. You can tension the spoke to specific kg force and then take a reading with the tensionmeter.
The Parktool meter is really sensitive to how eagerly you let the handle . If you let it slowly you can get considerably higher reading than if you let it spring back right away. So it requires a pretty consistent hand which not everyone has .)
The pro lock nipples are harder to turn but it’s not a big deal, they can be turned with a classic outside spoke key. I liked the squorx DT Swiss nipples more but experienced their original threadlocker drying too much resulting in issues in retruing the wheels later on. In my previous place I was eventually using unprepped squorx nipples with a violet loctite threadlocker as a finishing touch. I alway lace the wheels from inside the rim, it’s easier to spin the part of the nipple that contacts the rim rather then from the outside. You also leave no marks on the nipples which was an issue with the squorx ones because the black brass finish was a bit weaker compared to sapim. We have a square drill adapter for the initial lace up and then I use the Parktool SW-15 internal nipple tool.
Cool, thanks @Max_Odraska for the super detailed reply. Good to know that the park tool readings depend on how quickly you let the handle go. I wonder what speed of letting it go would produce the correct reading … In any case that would for sure be a way for me to under tension the wheel.
From what I gather so far, the best way forward for me would be to:
Bring the wheel back up to the highest possible tension that the manufacturer allows
Have the spoke tension checked by my LBS
If the tension loss persists, use either Sapim secure lock nipples and/or some Loctite 222
Also good to know that my combination of number of spokes, spoke type, my weight and my type of riding (mostly road with some gravel), does not seem to be doomed to fail from the get go.
The Park tension meter is not accurate and the conversion chart is useless when using bladed spokes. But, it is repeatable. So you can reliably use it to compare tensions in spokes on a particular wheel and to compare between wheels using the same spokes. So if you have a good wheel you can compare to that. Just don’t expect it to tell you what the actual tensions are, cxrays especially.
You do need to use it carefully, measure repeatedly and in a consistent position on each spoke. But it’s a useful tool for the amateur and genuinely accurate alternatives are very expensive.
Spoke fixing and locking nipples are only a compensatory measure for a fundamental problem of hub geometry though. Or insufficient tension overall but by the sound of things it’s the hub in your situation.
I have built 3 rear wheels with different spokes NDS:DS to improve tension balance. It’s the total cross sectional area of spoke steel that matters (2:1 lacing was popular on factory wheels for the same reason).
I suspect, assuming your tension measurements are close to accurate, that you will need to get the NDS up in to the 80+ range for it to be stable long term.
A cheap way to do this would be to buy unbutted DT Champion spokes for the DS and keep the CX-rays on the NDS. The wheel ride quality won’t be as nice but it will be stiff and solid if you are a bigger rider or like to sprint.
Thanks for all the useful insight, super interesting
I’m wondering why the conversion chart would be useless for bladed spokes and cx rays in particular. In principle you should be able to work out how a spoke bends regardless of whether is round or bladed. Granted the calculation might be a bit more complicated (I’m getting slight nightmare flash backs from my mechanics 101 class from studying physics), but it should be doable, no?
Regarding the position of where you measure, I have noticed that the readings change depending on where I measure. Is there a point where you’d recommend to measure or is it just about being consistent and then figuring out some actual conversion factor to get the absolute tension numbers? Now thinking about it, intuitively I’d say in the middle between the nipple and the last spoke crossing to have the least amount of influence of the boundaries. I guess I’d ask my LBS to measure with their tension meter and then compare.
By saying it’s the hub in my situation you mean that the resulting DS:NDS tension ratio is too extreme (100%:52%) and so my NDS will always be too low to effectively hold for a long time?
I just played around with The Spoke Length Project - Wheelbuilding calculator, specifically with the asymmetry parameter. Looks like with an offset of 4mm, I’d get a tension ratio of 100%:73% instead of the 100%:52% so I guess an offset rim would be a lot better. Are there any companies who make offset carbon rims like that?
