Hey folks,

I’m building up two sets of wheels for the new hardtail. First set is a pair of 650B+ – Spank Oozy 395+ on Hope Pro4 with DT Swiss Competition 1.8 spokes.

My normal approach is to lace, initial tension, then stress relieve, true, stress relieve, check that none of the spokes are pinging a really high note; and repeat until I’m happy that the wheel is true (which can take a very long time, depending on mood). I’ve never normally bothered too much about ensuring even tension across the spokes…

This time, after three cycles of truing, I thought I’d break out the spoke tensiometer and the Park Tools Wheel Tension App and try being disciplined. I evened everything up (using the figures on the spank site – so just under 120kgf on the rotor side, and about 105kgf on the other side – it’s a front wheel) and it completely threw the wheel out of true (about 2mm vertically and around 3-4mm horizontally). I re-trued it (with stress relieving) and the tensions were all over the place.

Surely this shouldn’t happen with a new rim? Do I just need to iterate again and again?

What’s *your* approach?

My original technique was pretty much the same as yours.

I now build with a tension meter from scratch and bring it to full tension. I found smaller, more frequent tightening passes around the wheel helped keep tensions far more even as they increased. At tension I stress releived heavily (treading on spokes), then checked and evened tensions, then trued, which was fairly minimal.

I’ve only built two pairs like this, but they’re far more stable in so far as I’ve not had to revisit them or re-true after a bit of use.

3mm isn’t considered to be a huge misalignment in a new frame (0mm is better, of course).

I’d probably get it true, then see if I can even out the tension through increasing/decreasing the tension on opposite sides. A bit of movement is probably preferable to uneven tension on a disc braked bike.

It’s possible of course to build an eccentric wheel with totally even spoke tensions and a perfectly round wheel – i.e. vertical lack of trueness.

Exactly what scienceofficer said.

It depends how stiff the rim is as to how much you can get away with, as youve just found a small mtb rim will still be true even with really bad tensioning.

Sometimes if a rim is really soft or for stans with their low spoke tensions I start with the “short” side (disk front, drive side rear) and get the tension even at arround 50% and the rim true vertically, then go round the other side a quater turn at a time, each pass only tighten the slackest 50% of the spokes (just be aware that if you tighten spoke 1 then you effectively add some tension to the opposite side as well (arround 16) so avoid just tensioning half the wheel if you do this).

It takes a bit more practice, but it is possible to get the wheels perfectly true and the tension within about 5% (+/-1 on the park guage). The payoff is much more stable wheels.

The wheels on my sscx take an absolute hammering on byways and bridleways and are still perfectly true inspite of that.

I am using Stan’s Flow Mk3s FYI. As well as demanding lower spoke tensions, they are meant to be a bit more compliant vertically (hence the flatter section). I had this in mind when a bargain on an x-tools tension meter cropped up, and I was thinking about upscaling my building techniques anyway.

From my experience of building 3 mtb wheelset (2×26 and 1x650b) and then recent rebuilding with new spokes of a 24 spoke road rear wheel it’s all about trying to get the best compromise.

The first point is about you tension meter – you can’t rely on the exact tension is says the wheel is –
due to variance in the calibration of it and how you measure tension on the spokes. What it does do is help you to achieve even spoke tension across each side of your wheels.

I’ve found completely even tension doesn’t always give you an immaculately true wheel – so I’ve to get them as true as I can with broadly even tensions. So far the 3 sets of mtb wheels haven’t need revisiting to true or tension so I’m hopeful my approach to it has worked.

With the build I’ve aimed to get the wheel laced and then all the spokes nipped up so all of them are just enough I’m the nipples so that there is no thread showing.

I’ve then looked at vertical and lateral true and tweaked things so that’s broadly correct but not at finished levels or horizontal true. I’ve then gone round on the higher tension side (so disc side front and cassette side rear) and worked on the spoke tensions and tried to get them reasonably equal. I then look at trueness again and try to make adjustments to the other side of the wheel to achieve this.

I’d then have a look at the dish to check it’s broadly correct. Once that’s done I’d ease the spokes – having watched a video of a top Dh wheel builder I put thick cardboard down on the floor and stand on all the spoke pairs in turns until the wheel stops pinging.

I’d then retrue and work round all the spoke tensions again.

I work through this round by round until I’m happy the wheel is broadly true, Ty ending is correct and all the spoke tensions are in the same ballpark.

So far so good – although I’ve only ridden the rear road wheel I’ve rebuilt 3 times so far. I’m more nervous about that as the tension on the drive side appears to be over what the manufacturer recommends for the dish to be correct, but I can’t wind the tension down on the non-drive side much without spoke threads showing. I wonder if the non drive side spokes are too long, but I’ve used the correct length spokes and nipples as advised by the manufacturer so I figure my tension meter must be over reading tensions a bit (it’s also the Park Tools cheapest tension meter).

Joe makes a good point about tension meters. I’m fully aware that mine is cheap and may not report the exact tension, but, used correctly on the same point on each spoke, I will get commonality of reference across all spokes, but it’s important to use a technique that gives you reproduceability.

It’s easy to see how important this is if you test the spoke tension up and down it’s length, you get quite different readings.

Scienceofficer makes a good point…

if you test the spoke tension up and down it’s length, you get quite different readings

Spoke Tension ‘Meters’ don’t measure spoke tension directly.

I wouldn’t get too hung up on the actual tension. I’d worry more about relative tension between spokes and between sides of the finished wheel. I tap each spoke and listen to the sound – I think that gives a good enough guide.

Also spoke material and spoke thickness.

a point comes where the rim is as round and as straight as it wants to be, and trying to be rounder and straighter will likely involve squiffy tensions

I work on getting even tensions to within 5% – getting it lower often requires substantially more twiddling. It’s easier with stiffer rims.

So, this is a pretty solid rim – I think it’s had the least response to small adjustments of any rim that I’ve ever worked with (it’s an i35 32H, so there is a lot of material between each spoke). That made it all the more surprising that it deflected so much.

I’ll try going through a few iterations of the process of tensioning, truing, stress relieving. I will try the “stand on it” method (I’ve always used my hands or an old crank, pushed into the crossings and twisted). I think I’ll give @joebristol’s method a try – even up all the tensions (probably on the low-tension, non-rotor side),

@dovebiker: interestingly, the Park Tools app defaults to 20%!

I should also say: thanks everyone for your thoughts!

Some pretty brutal methods of stress relief up there, what are they actually supposed to achieve? Aren’t we just taking tension off the spokes to release any wind up?

Are you sure you haven’t compression buckled the rim? Before now I’ve added a uniform extra half turn on all the spokes of a wheel and gone from within a mm of true to easily 4-5mm out. Backing off a bit of tension and the rim came back towards true again. My conclusion was that I’d just pulled too much compression into the rim and taken it past it’s compression bulking limit and if you’ve got beyond the buckling limit you’ll have a nightmare trying to hold it true.

I know you’ve got the numbers from the spoke meter and compared those with the recommended values, but if the reader isn’t accurate could you be a bit high? It should be a pretty robust rim though at that size which might buckling unlikely.

@13thfloormonk

No, the primary function is not to release windup. However, is not necessary to be brutal.
On both points see this link

Wheel Building Tip No. 4 – How to Pre-stress Your Wheel

@mikeys – no, and I totally accept that the accuracy of the meter is likely questionable. Still, it would have to be reading *very* low, I think, and my secondary tensiometer (plucking and my ear) suggests that I’m on the low end of taut enough. I’ll try a tone-comparison against another 650B wheel.

The Spank FAQ suggests building the wheels on the looser side, though – I’ll try taking a half-turn off every spoke and see where that leaves me.

I saw a video by mavic or dt Swiss or similar and they squeeze parallel pairs of spokes and then lay the assembly in their laps with their legs supporting the rim about 1/4 of the way in so that 1.30 and 4.30 o’clock and 7.30 and 10.30 are touching the thighs. They put a towel on too of the wheel. Then they put their elbows on the rim at about 4 and 8 and grip at 10 and 2 and push down with elbows and pull up with hands. Turn the wheel and repeat over and over until you’ve done the wheel about 6 equally spaced times. Not brutal, nothing bends but their is some flexing and popping.

I’ve always been surprised by how much fitting a tight tubeless tyre affects spoke tension, I always re-tension after

The problem is likely that you’re creating different effective spoke lengths before tension is applied through slightly different nipple winding on, when you tighten up you end up with spokes of different lengths as their nipples are wound on or off by +- say 1mm, which when applied equal tension too will cause the wheel to be off true. Think of the truing being a case of centering the hub in the rim rather than the other way round and its easier to visualise how this can be a problem. We handbuild maybe 40% of the wheels we sell – its a simple process – we start with a known spoke/rim combo. We use a DT nipple driver to apply the nipples to a preset length to each side of the wheel so we end up with very equal length with modest tension. The resulting wheel is usually within 1-2mm of true and round without any real work. Then we add the final tension (4 half turns), check for outlier spokes, and assuming the tensions are even then the wheel should be within 1mm now. Then we give it a final true to finish it off using very small adjustments in pairs of spokes. Even on the worst wheels (+- 2mm out say) we still go round and round making small corrections rather than trying to fix an issue in one go. Finally we recheck the tension, its normally within +- 2% of the target, which we consider to be fine. Between each step we relieve stress.

A pair of ready to ride wheels takes around 50 mins from start to end, including tape and finishing up. The key though is the ability to deliver very equal wind on the first nipple tightening such that your 80% tension wheel is already rideable. If you can get to that the rest is just tidying it up.

I think doing smaller tightening increments allows me catch the discrepancy in effective length earlier in the jig and thus minimise it as I go. Net effect similar, but less precise than Ben. Also probably why it takes me an hour to do one wheel!