Tensiometer calibration
 

There has been previous discussions about fixtures and home grown methods for doing this. Now that I am looking at ramping up my wheel building business, I decided I should take steps to monitor my good old Wheelsmith meter to ensure reliability.
Lots of good ideas were tossed about in this thread some time ago.
http://www.bikeforums.net/showthread...ration-fixture
Anyhow, I went ahead and built my own fixture which seems to work very well. One thing I learned was that in the 9 years or so that I've had the Wheelsmith unit, the calibration remains pretty much spot on.
Here's a video I made showing the purpose of, and finally the construction and functioning of the calibration fixture.

https://www.youtube.com/watch?v=xgsz7l1GWoI

FYI, you might want to use much accurate tools to check your wheels. Wheelsmith and Parktool are okay for wheel truing but not for wheelbuilders. DT Tensio or Wheelfanatyk is recommended. I've been using my Tensio for some time now and have been very reliable.

Well, there may be something to be said for having bearings on all three contact points, but I'm not sure I buy the argument that those, or using a dial indicator in place of a graph take the accuracy to that much higher degree.
I would be interested to hear what the expert's take on that would be, and am willing to be convinced otherwise.
I guess the question is, just how precise does it have to be to fit within the tolerance of acceptable tension?

Tensionmeters like all measuring tools provide an "estimate" per object measured.

And of course, when we build a wheel - we are taking an "average" of the "estimates" as an overall read or statement about the wheel.

Dials, digital readouts and the like are cool and all that - but unnecessary for the task of building a wheel. The resolution they offer are not required of bicycle wheels. They do come in handy with a lathe though...

The real difference between cheap and expensive tensionmeters is more than anything else about calibration period.

The Park is a very low cost item - but requires frequent calibration. A volume shop that wants each builder focused on putting out wheels isn't going to want those slowing down their builder every time the builder has a reason to suspect the tool needs re-calibration. Other than that - it works just fine. I have two...

The FSA and DT are very high cost - but you can bet they will go long periods of time before a re-calibration is needed - ideal for the shop environment noted above. But cost-wise, not justifiable for the low volume or home builder.

=8-)

I'll stick with my park tensiometer, but I do like the idea of building a calibration method. Sounds like a fun project. Maybe even as simple as hanging a known weight on a suspended spoke and comparing the expected deflection against the tool's measurement.

I've got a spoke hanging from the ceiling, and a bosun's chair that hooks on to the spoke. I'm the known weight, I get in to the chair, hook it up and I'm all set for a one-point calibration control. If I want another point, I'll recruit either a skinnier or a chunkier relative, neighbor or riding buddy. Any one ready to be honest about their weight really.

So if you only want to calibrate using one weight, would it be useful to use various spokes to test the differing deflections for these spokes with a given tension and thus testing various ranges of the tool?

Once again, as stated in previous threads:

A tensionmeter only sees a given amount of deflection for a given distance.

It does not see anything else, including the guage, butting, or brand.

It is a chart that does the interpretation - guage per guage...

If there is a range - it's the interpretation - not the tool.

=8-)

Um... yes... which is why I would use different spokes to verify the accuracy of the tool against its included chart across a range of total deflections.

I'm curious. To those that are re-calibrating your gauges, how much variation are you experiencing? Especially, if you have much experience with the Park TM-1's.

Secondly, unless that variation is considerable, is it really of much concern?

My primary reason for initial use of a tension meter was to ensure even tension of spokes around a wheel and calibrate my fingers to what they were feeling. I know some like mrrabbit will have the finely calibrated fingers and ear to evenly tension a wheel by feel. But, my meter still suggests that it is more precise than my fingers or ear in this regard.

It seems that calibration wouldn't really be pertinent to tension equalization.

My second reason is to ensure I know to what absolute tension the wheel is built. Here I see the value to having a calibrated meter/interpretation card. But, a single point calibration isn't going to verify the meter against the range of the interpretation method. The first issue I see is the assumption that the spring rate will first be linear and then that it will remain so over its life. Any change in the spring rate over its range of travel and a multi point calibration for each gauge would be neccessary for modifying the interpretation chart.

If the spring rate is linear and isn't changing over the course of its life, then why can't we just zero the meter and know that the deflection will be accurate for the predetermined interpretation chart?

Whole point of re-calibration is to bring the tool back into sync with the chart so to speak. Once again, it's a rough process - but quite okay as we don't need the precision or resolution necessary that one might need when doing machinist work such as turning a piece on a lathe in the thousandths of inches.

Of course, ideally, one would want to calibrate guage by guage - and redo the chart for each.

If you have the original Wheelsmith chart, you'll notice that they are referencing their own Wheelsmith brand of spokes quage by guage - not others.

Like many brands with different level of spokes - CN has the standard and Mac spoke. The standard is slightly lighter for 14g than the Mac - and behaves a tad different deflection wise. An ideal chart would accommodate that - same for Pillar (Richman), Shun Lih (SL), etc.

But once again, our wheels are are tensioned to rough numbers - 110 kgf for example - not 110.00.

As to the Park Tool TM-1 - the reading for 14g will start to over read by about 5-7 kgf after anywhere from a dozen to a hundred wheels. So I end up having to tighten up the spring. For my first of two, I'm about ready to email Park and ask, "Do I request a new spring, or would you guys like to have the whole tool for testimonial purposes?"

The constant re-calibration doesn't bother me - the tool works just fine.

=8-)

I'm pretty new at the wheel building game, so I'm keeping to relatively "generic" 32/36 spoke wheels.
The consensus seems to be 110Kgf on DS rear and 90ish for front.
IF my tension meter is accurate @ 100 Kgf, that's plenty good for me. I don't care about 60 or 160 Kgf.
At least not yet.

Yeh,

And, my limited experience suggests that spokes vary sufficiently from one brand to the next or between lots to render the interpretation chart more or less inaccurate. I don't have my micrometers in the garage, but, when I can get one home I'll mic a few spokes and see what we get.

I've also found that the chart supplied with my TM-1 does not yield the same result as entering the meter value into parks TCC tension calculating spreadsheet. Which I could reconcile with a calibration check.

Which still leaves the question of "just how important is absolute tension?" somewhat unanswered.

Basically, I'm learning that if I build wheels that are well tension balanced they hold up exceptionally well (so far). Becasue of my weight I have been hedging toward the top of the manufactures tension range for my rims. Or, even a little over, +10kgf. And given the possibility that my meter may not have matched the conversion chart to begin with, may have altered since then and that the chart doesn't match the calculator I may be off by 5-7kgf. (probably under)

rabbit, do you find that your tool stabilizes at the +5-7kgf, or, does it continue to drift if left uncorrected?

And, when you say it 'works just fine', do you find that you need to take several readings on each spoke to ensure that any spurious readings are discarded?

I've found that even with frequently application of nose grease to the indicator slide slot I get the occassion 'off' reading. It seems as though I sometimes need to read the same spoke several times to get a accurate indication of what the tension is. On your everyday 1.8mm spoke a variation of 1/2 a gradation on the deflection scale is equivilent to 7kgf. And, I would say that I'm not unaccustomed to seeing my readings on a given spoke vary by +/- 1/4-1/2.

Those sound like pretty common numbers. I run slightly higher DS tension at 130kgf to increase my NDS tension while maintain correct dish and eliminating the issue of NDS spokes backing off without the use of loc-tite or anything beyond the lightest of coats of spoke prep. At 110kgf I was finding that accute road hits (rocks & pot holes) or standing climbs could result in lateral deflection of the rim and one or more NDS spokes loosening. But, I'm a BIG boy.

Remember once again, our tensionmeters are estimators - and support us in arriving at a ballpark so to speak. Remember also that we are taking our estimates of each spoke and averaging them for the wheel as well. In other words - you really should step back from trying to be precise or accurate - that's not the purpose of the tool. Most folks are aiming for roughly a 10% percent margin in relative tension AND roughly the same between per spoke tension and the desired goal.

Once again as an example- we are aiming for approximately 110 kgf - not 110.00 kgf. And I'll try to get most spokes in the 107 to 115 range (24-24.5) - not 107.00 to 115.00. There's no point to it - an estimate is all we need.

Here's what I do with the Park TM-1:

1. It floats off a little at a time - each time I'll find my brain in the background saying, "Wake up host! That doesn't feel right..." More often than not, checking the tool against my calibration rod or wheel for that specific tool when first used will confirm that it's off by about .5 on the dial.

2. As to readings it is very consistent. Soft release will show the same reading 9/10 times. Hard release with tap and settle will show same reading 10/10 times. Not everyone's soft release is consistent. Sometimes we slip.

When I label my calibration wheel and rod for each tool - I record both the consistent soft release result AND the hard release with tap and settle the first time the tool comes out of the box.

The Wheelsmith reads the same whether soft or hard release - better tool but with less resolution. Black and white in operation - no quibble.

3. You should not need to lubricate the TM-1. If you find you need to do that - you need to check the tightness of the main screw/bolt holding it together. I have it such that it just minimizes side play while still allowing free non-stick movement. If you can't resolve this - maybe you have grounds for an RMA with Park.

In my opinion, when a measuring tool's operation leaves room for doubt as far as mechanical consistency is concerned - you have a valid reason not to trust it. I'm betting most here would say that's a reason to discuss RMA with the manufacturer. I would be inclined to do so myself.

=8-)

Very clever, Dan! Thanks for sharing this creative and practical way to check calibration.

Have you tried using the next thinner gauge spoke on the NDS to give it "relatively" higher tension? (slightly more "stretched")
Some use a straight 14 ga. on the DS and a 14/15/14 DB on the NDS.

I have. But you're not achieving "higher tension", just greater elongation of the spoke. The tension is dependent upon the DS tension and bracing angle regardless of NDS spoke diameter. It's simply that in achieving the same tension you are causing greater elastic elongation of the thinner spoke.

And how does "greater elongation" not result in higher tension? Any time you tighten a spoke nipple it stretches (elongates) the spoke, and the result is always higher tension.

If we had been discussing additional tightening of the same spoke you wouldn't be wrong. But, not when the "greater elongation" is the result of the decreased cross sectional area of a narrower gauge spoke, as suggested by Bill.

I've built with with 2.0 both sides, 2.0DS/1.8NDS and 1.8 both sides. The amount of tension required of the NDS spokes to pull a wheel into correct dish does not depend upon the gauge of the NDS spokes. As previously stated that tension is determined by the DS tension and the bracing anglee of both sides. To achieve the requisite tension a thinner spoke will have stretched more than a thicker spoke.

Does that make sense to you? If not, I'll suggest a metallurgy text or two.

Elongation (aka strain) is a function of the stress in a spoke. The stress is related to two things: tension and cross-sectional area. Tension is a force, cross-sectional area is obviously related to the diameter. When you tighten the spoke, you force it to elongate by the amount you've pulled through the spoke nipple, and the tension increases in proportion with that. For a given spoke, more elongation will mean more tension, but for several different spokes of various different cross-sectional areas, the same elongation (i.e. the same number of turns of a spoke nipple) will produce lower tensions for spokes with less cross-sectional area.

And the greater elongation means the NDS spoke is less likely to back off.