Road Bike, Cycling Forums banner

Spoke Tensiometer vs. Frequency Measurement

16K views 33 replies 12 participants last post by  fallzboater 
#1 ·
I've built myself a number of wheels over the years with good results, but I haven't had access to a tensiometer or other method of measuring the actual spoke tension. I recently bought my daughter an inexpensive Korg tuner for tuning her flute and would like to use that to help achieve the rim manufacturers recommended tension and equalize tension among spokes. Starting with the information and formula linked from http://www.bikexprt.com/bicycle/tension.htm I wrote a simple spreadsheet that allows me to input spoke length, diameter, wire density, and desired tension and give the estimated frequency, or vice-versa. For example, 284x1.8mm spokes at a tension of 130 kgf would theoretically give a frequency of 440 Hz (A4 note).

With thin, straight-gage spokes on a radial-spoked wheel, I would expect the formula to work well, but it'd be interesting to check it. I'm not sure how much error is likely to be introduced by either butted or crossed spokes, but with some testing, a correlation could be made. I could do this myself if I had access to a tensiometer and a few different types of wheels, but I don't. Has anyone with a tensiometer and a tuner either tried this, or would be interested in testing it out? I can send out the spreadsheet by e-mail if you like.

-David
 
#3 ·
You can....

BUT.


This is not a simple math problem...amongst other issues:
-Spokes are not ideal, usually spokes are butted-which complicates things greatly
-The audible pitch is not a "pure" note it has overtones, and as such tuners easily grab and listen to the wrong overtone.
-For a wheel to be round and true, a well built wheel will usually have +/-10% range in tension which is quite a range of pitch...you can use pitch if you have a good ear to hear if the tension if fairly close though.

Thus in building a wheel-you want a tensionmeter--pitch by itself is not reliable enough


If you want to go and calculate values for an ideal wire (string:

f = √(TL/m)/2L
where

f is the frequency in Hertz (Hz)
T is the string tension in Newtons (N)
L is the length of the string in meters (m)
m is the mass of the string in kilograms (kg)
√(TL/m) is the square root of T times L divided by m
 
#4 ·
We experimented with this. The short answer is ... it doesn't work. Not only could we not generate an algorithm to give tuning pitch versus tension, but we also asked a couple highly reputed European builders to do it, both of whom had experience with audible tuning. They couldn't even generate as consistent a spoke-to-spoke tension as an amateur with a tensiometer. They (and we) concluded that too many issues affect audible spoke tensioning. We noticed people had a lot of trouble with harmonics and with secondary vibrations (not strictly harmonics) on the same spoke and on overlapping ones. Basically the stretch you flick is the stretch that you want to hear vibrate. But you'll generate vibrations at a completely different frequency in each of the other sections of the spoke that are separated by spoke overlaps, plus you'll generate lesser vibrations in the spokes the cross over the spoke you're flicking. We tried different flicking instruments (fingers, guitar pick, a tempered steel tool, and even a modified harpsichord plectrum that would actually pluck the spoke and theoretically be the most consistent). Nothing worked all that well. The amateur with the tensiometer beat a professional every time.
 
#5 ·
Yeah, that's the formula I used. My daughter's tuner is at school right now, so I haven't tested it. I was thinking I could take the tuner to my LBS and they'd probably let me test some wheels against their tensiometer.

It appears that a 10% increase in tension would give an ~5% increase in fundamental frequency.

I'm using a set of new Velocity Deep V rims, and I'm having some trouble getting them round within ~1mm with even tension, especially near the joint. I'll have to decide what balance I want between even tension, and roundness. I've done a number of wheels with eyeletted Mavic rims (like Open Pros) and they seemed easier to to get round and true with even tension.
 
#6 ·
fallzboater said:
Yeah, that's the formula I used. My daughter's tuner is at school right now, so I haven't tested it. I was thinking I could take the tuner to my LBS and they'd probably let me test some wheels against their tensiometer.

It appears that a 10% increase in tension would give an ~5% increase in fundamental frequency.

I'm using a set of new Velocity Deep V rims, and I'm having some trouble getting them round within ~1mm with even tension, especially near the joint. I'll have to decide what balance I want between even tension, and roundness. I've done a number of wheels with eyeletted Mavic rims (like Open Pros) and they seemed easier to to get round and true with even tension.
With realistic equipment-you should be having difficulty.

Every rim is different in how it tensions up to round and true....only if one has ideal rims with ideal spokes, ideal hubs and ideal spoke nipples--would even tension result in a round and true wheel. If your wheel is round and true, and you're +/-10% tension, then you've done your job.
 
#7 ·
11.4 said:
We experimented with this. The short answer is ... it doesn't work.
Thanks for the reply, although that's not what I was hoping to hear. If I built more wheels I'd get a tensiometer, but I do so few these days that I can't really warrant it (plus I build pretty good wheels anyway). I was hoping the tuner would at least work well for balancing tension, even if not for determining the precise value. I'll still give it a shot just to satisfy my curiosity. I was thinking that holding (damping) the crossing spoke might help, and it also seems that you might need to use the free length between the crossing and nipple, not the actual spoke length.
 
#10 ·
danl1 said:
Something I stumbled across on my interweb journeys:

http://www.bikexprt.com/bicycle/tension.htm
Yeah, that and Sheldon Brown's site were the two I found initially on this subject.

As a follow-up, I just got an e-mail reply from Velocity regarding recommended spoke tension for the Deep V rims.
"We use and recommend the following for a Deep V.
314.7 pounds / 1400 Newtons / 142.9 kgs "

With a typical spoke length for 3x lacing (drive side CK rear hub, 280mm), that would correspond to the following fundamental frequencies:
2.0mm 421 Hz
1.8mm 468 Hz
1.6mm 526 Hz
1.5mm 561 Hz
 
#11 ·
Many people are tone-deaf. With tension-meters available today, learning to build wheels really should be done with them. It
is simply more accurate. After you have building down, then employing tonal frequencies can be experimented with as a hobby.

Of all the spoke tension-meters out there, I'd highly recommend the FSA - designed and approved by Jobst Brandt. Author of "The Bicycle Wheel" - the first, and still the best book, on the physics of a wheel. They will run around $250. Failing finding one (they are scarce and are a labor of love, rather then profit) of these, the Park Tool TM-1. Around $60. Which is about as accurate as the Hozan $380ish, or the DT model - same or over $500 as they have two models.
 
#12 ·
i have the tm-1, and i picked it up for $50 on ebay. I fail to see how it is expensive?? It reads spokes consistently, and is a pretty good guide. There's no point to be precise with balance, since u end up adjusting the spokes at final true anyway... its a great tool, and i probably would not have embarked upon rebuilding/building wheels without it.
 
#13 ·
Great! It is a very fine addition to the bike-mechanics armory. I could go into the physics of why it not as accurate - But I won't bother*. The inaccuracy is negligible for common usage.

*It would bore people to tears.
 
#14 ·
I did find a TM-1 to borrow, and now that I've used it, I would never try to finish a wheel without one. Although I could build a tight and true wheel without it, I found that adjacent spokes one one side of the wheel could differ in tension significantly (I'm not sensitive enough to the tones to be able to tell.). With the TM-1, I was pretty easily able to keep the tensions on each side of the wheels to within +/- ~5% of the average value. I expect this will result in a wheel that will require less fiddling later on.

-David
 
#15 ·
I can get extremely consistent results by equating tension to pitch. I'd like to think I have real good ears having been a musician for a couple of decades. The one thing you can't count on is using pitch to measure tension. I measure tension with a tensiometer and then use pitch to make sure the tension is even around the wheel by having a consistent pitch.

A tuner will not work. The pitch has too many overtones and will not ring long enough for a proper reading. I don't recommend the pitch method to others simply because I don't know how good their ear is. I can only vouch for mine and they are on the money with the machine.

Each wheel is different because of the spoke length differences and the cross patterns. The cross patterns will give you wild overtones. Radial is of course easier to hear.

Bottom line. If your are good, you can make sure the tension is even with pitch, but don't rely on it for measuring the tension. Use a tensiometer. I understand if people are only occasional builders, but something like a Park is pretty inexpensive and worth it.

-Eric
 
#17 ·
Ichijin said:
Wouldn't a tuner which uses vibration instead of sound work better?

I have seen one for guitars which was like that
Nope. Same problem...and a few others are added as well.
 
#19 ·
Puchnuts said:
The frequency, if all the factors were entered into a decent computer, would tell you the tone. As would the tone tell you the frequency. Now we can wait for someone to write the conversion software.
There are straight look-up tables available to give you the frequency of different music notes (tones), or am I missing something?

The formula to convert tension of a wire to fundamental frequency is quite simple and can be plugged into a programable calculator or spreadsheet. I think for radial lacing of straight gage spokes, a simple electronic tuner might work OK, but I haven't had much luck with the Korg tuner and my 3x wheels. Maybe it'd work better if you actually performed a spectral analysis of the tone, but that's getting a lot more complicated than I was interested in pursuing.

Good replies in this thread. I'm not a musician, and after borrowing a TM-1, found that I was not good at equalizing tension by either nipple torque or tone. I found another buddy to go in on a new TM-1 with me, so I plan to use that to tune up our existing wheels and for all future builds. I'm still a big fan of Brandt's building techniques, although I'm not sure about his method of determining maximum tension, but I think with his book, a TM-1, and some practice, just about anybody can build wheels comparable to the pros.

I understand the sources of innacuracy of the TM-1, but I don't believe the absolute tension value is as important as even tension. I get good repeatability and I believe the precision is adequate, especially since you'll normally need to deviate a bit from even tension to get the finished wheel round and true enough (maybe not with disk wheels).

-David
 
#20 ·
Having perfectly equal tension will get you an out-of-true wheel. Having a perfectly true/round wheel will give you unequal tension. Jobst Brandt is of the school that equal tension trumps being in true/round. Others feel differently. I aim for getting as close as possible to both. While still being realistic.
 
#21 ·
I find using tone a very good measure, but only for equalness of tension. However, I do believe that absolute tension is overrated - if you are within the design limits of the spokes/rim within 10%, you will be fine. Just consider the increase/decrease in spoke tension if you hit a pothole or smth. similar at speed - I'd guessimate at least a doubling of individual spoke loads!

Anything that gives you a snakebite on a properly inflated tyre probably produces peak spoke loads of 300% normal or more.

Crossing spokes actually are easier than radial ones to judge equal tension on: If the 2 spokes are at equal tension, both spokes will ring true when you pluck only one of them. The more unequal they are, the more of a 'thud' sound you will get.

An equally tensioned non-dished (front wheel, or assymetrical rear rim) wheel actually will ring as a whole when you strike the tire with a screwdriver handle.

I don't use a tuner, but rather my flute. So I pick a note that seems to be close to the desired tension, let's say an A or a B, and use that to tension the spokes. I do this simultaneously while truing, because, with a little thinking, you can optimize both things at the same time.
 
#22 · (Edited)
Sriajuda said:
I find using tone a very good measure, but only for equalness of tension. However, I do believe that absolute tension is overrated - if you are within the design limits of the spokes/rim within 10%, you will be fine. Just consider the increase/decrease in spoke tension if you hit a pothole or smth. similar at speed - I'd guessimate at least a doubling of individual spoke loads!

Anything that gives you a snakebite on a properly inflated tyre probably produces peak spoke loads of 300% normal or more.
Radial deflection of the rim results in a decrease in tension of the lower spokes, and only a slight increase in tension of the other spokes. When you can run into trouble is when the loads cause spokes to go slack, not too tight. Higher initial tension (and a radially stiffer rim) means a stronger wheel. I'd be surprised if the peak spoke tension ever approaches 150% of the static initial tension (assuming you're starting at 100+ kgf).

I think getting within 10% of a desired average tension level is great, but difficult to do if you don't have a tensiometer or a similar known good reference wheel (or even if you do). You musicians have an advantage over most of us for equalizing tension, I'm sure.
 
#23 ·
fallzboater said:
Radial deflection of the rim results in a decrease in tension of the lower spokes, and only a slight increase in tension of the other spokes. When you can run into trouble is when the loads cause spokes to go slack, not too tight. Higher initial tension (and a radially stiffer rim) means a stronger wheel. I'd be surprised if the peak spoke tension ever approaches 150% of the static initial tension (assuming you're starting at 100+ kgf).

I think getting within 10% of a desired average tension level is great, but difficult to do if you don't have a tensiometer or a similar known good reference wheel (or even if you do). You musicians have an advantage over most of us for equalizing tension, I'm sure.
You are right in stating that radial deflection causes the lower spokes to slacken. However, just proportionally the upper spokes get a much higher loading, which is what breaks spokes (apart from obvious things as too loose spokes in general, and highly uneven tension amongst the spokes).

Of course, initial loading at 120-150 kgs is pretty high. But take a 80 kg rider plus 10 kgs of bike, hitting a quite moderate ridge of, say 3 cms height at 35 km/h. If the rider does not 'lighten up' (essentially allowing the bike to pivot around the cranks), the hit especially on the rear wheel is immense.

I very much doubt 150% of initial load will be remotely close to the actual spike loads. Under these conditions, even the stiffest of rims (discounting the strength by the aggregate of spokes) will yield like pudding.

10% difference in frequency is *very* detectable even to the untrained ear. I'm just a dabbling amateur musician, and I can easily hear frequency shifts that exceed 2 Hertz, sometimes less than that.

Twinging 2 spokes in succession it can be difficult to tell which of the 2 frequencies is lower that the other, surprisingly, although it is obvious they are not the same.
 
#24 · (Edited)
Sriajuda said:
You are right in stating that radial deflection causes the lower spokes to slacken. However, just proportionally the upper spokes get a much higher loading, which is what breaks spokes ...
I'll have to disagree with you, there. I think you'll find that if you load a wheel by having a fat friend sit on a bike, you'll find that the lower spoke tensions decrease to support his weight, but the upper spoke tensions don't increase to anywhere near the same degree (maybe not even measurably). The load is not shared between the upper and lower spokes. Try it out.

I believe spokes break not from overloading, but in fatique where there is a stress concentration, such as at the elbow or at a notch caused by dropping a chain or other damage. With enough fatique cycles, the spoke will break even though the ultimate tensile strength has never been exceeded (in fact, for a given alternating load, fatigue damage is worse if the static load is lower). Statically loaded swaged spokes can yield 5+mm before failing in tension, and that's not going to happen before total rim failure. If the wheel has been properly built and stress-relieved, stresses should remain below the fatigue limit and spoke life is theoretically infinite. That's why you should be able to continue to swap rims as they are damaged or crack, using the same spokes (as many of us do), and never have any failures. Note that aluminum rims do not have a fatigue limit, so they can be expected to eventually crack around the spoke holes or eyelets, no matter how well built.
 
#25 ·
Sriajuda said:
You are right in stating that radial deflection causes the lower spokes to slacken. However, just proportionally the upper spokes get a much higher loading, which is what breaks spokes (apart from obvious things as too loose spokes in general, and highly uneven tension amongst the spokes).
No... not in a "normal" wheel anyway. The bottom of the rim is effected... near the contact point. The rim flexes towards the hub, which reduces tension in the bottom spokes and equalizes the forces... ie less tension in the lower spokes gives a net upward force at the hub. The upper spokes are effected very little.

I very much doubt 150% of initial load will be remotely close to the actual spike loads.
Not an increase due to radial loading, but lateral+torque loads can cause a 50kg+ increase.
 
#26 ·
rruff said:
Not an increase due to radial loading, but lateral+torque loads can cause a 50kg+ increase.
I don't know how you get to that. It seems like a high number just for the torque load. Maybe that could happen in low gears on steep climbs, but not for most of the miles you do.
Even if that's correct, the torque load will decrease the leading spokes as much as it increases the trailing spokes, and the rider's weight adds to the decrease as well. I have a fiinite element analysis that says that the maximum reduction on spoke tension from the riders weight is 35% of the load. That's for a 36 spoke wheel, an it would be more for a wheel with fewer spokes. So a 180 kg rider unloads the spoke by about 30 kgf, plus another 50 kgf from torque, that doesn't leave much tension or margin for error. A NDS spoke can be less than half the tension of the drive side, so that might be only 55 kgf or even less. Of course, you can't really quantify any of this stuff with a finite element analysis, which I don't have the capability of doing now.

em
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top