Loading

Wheelbuilding Tensiometer (develoopment)

Please Login to Comment

We mooved here because the extreme long comment section have to end.

So this should be a possibility to develoop a precise, cheap and of course printable Tensiometer like this: https://www.thingiverse.com/thing:2810089/#comment-1835147

Tensiometer

Sorry for being late to this party, I have printed and used the DT Swiss clone and I have parts ordered to build a calibration jig based on Dan Burkhart's (https://youtu.be/xgsz7l1GWoI) design. What I am doing different is using 2020 Aluminum Extrusion instead of welded steel tubing. I don't know how to weld and I don't have the equipment. I have several standard size compression springs that I will be testing for repeatability as well as ease of use. Right now I am printing @cmh's version as you cannot have too many tools. I do need to get a couple more dial indicators so I don't have to keep taking a gauge apart to move from one version to another.

By the way, here in the US I have a Harbor Freight near me so I bought this spring assortment, https://www.harborfreight.com/200-piece-assorted-spring-set-67562.html

Hey thanks for inform! I've seen your DT-design before and want to respons but dont't have the time until now. But I have to print it as soon as possible. Im not sure if this is the right place to discuss the development of a calibration rig (so maybe you guys into another special bicycle forum/thread for that!?) I know the video and other ideas for calibrationrigs. Anyway enclosed some pictures of what I made last weekend.
I decided to build it from materials which I still had (because I am out of money and want to try to keep it "low cost" as possible).

To bring everyone up to speed, we're talking about the Jobst Brandt design tensionmeter as he documented in his book "The Bicycle Wheel". These have been available for from various sellers, I've got an FSA model with a dial gauge, but that's unfortunately out of production. More recently, Ric Hjertberg of Wheel Fanatyk has been offering his version in both digital and dial gauges. Since the Jobst Brandt design is pretty simple, I was thinking this is a perfect application for 3D printing, but recently noticed that Modellaner totally beat me to the punch before I had done anything but collect reference material!

You can read our discussion up to this point in the comments of Modellaner's thing (linked above) but summing it up, I'm printing his design with slightly different parts, so we've been adjusting the parts. Concurrently, I've been working on a testing fixture for calibrating the tensionmeter, and we've also been discussing simplifying the design similar to the FSA or Wheel Fanatyk model, neither of which have the linear bearing.

So, that's where we are right now - my version is almost done and hopefully should be ready for testing soon!

I've done a reprint of the body - noticed that the bearings weren't turning freely as the screw was holding both parts of the bearing against the body. Probably not a huge issue, but easy enough to fix. Attaching the STL of my modified version. As I'm now down to just the one printer with the 120mm build volume, I had to trim the sides to make it fit. On the upside, it's now all red!

Still planning on using both tensionmeters to check the tension on a wheel so we can see how they compare, then building the calibration fixture.

That was a problem, I have had the same issues. I used for that the "thin M4 washers" as described in the description.
But your solution is the best I think! So I will upload the stl. file. I think it's a good idea to compare the original with the printed one, to make shure how exact it works. And I think that we have to change the size to the original one, so that a human with normal hands can use this.
The thing that load the spoke is actually to big (I noticed that because I build a 24" wheel and can't get through the spokes)
But those things we only can change if we change the whole design as well as possible to the original.
I would to do that, but as you know I dont have the original measurements. By the way I found http://s1023.photobucket.com/user/masterquang/library/3D%20Printed%20Tensiometer?sort=3&page=1 by searching for the measurements.
Seems that the work already to be done but I cant find a upload of this model. The designer say that it works but the carriage seems a little to thin and maybe will be unstable and bend away. I think we can improve that design, by use bearings on the edges and maybe a more stable carriage.

Whoa, shit, we need to get "masterquang" in here, looks like we're just duplicating efforts. I'll be studying those photos, and trying to get in touch with that person to see if we can't get those models.

One thing I'd request on any redesign is keeping the width under 120mm so I can print it on my little printer. Since the bearings only have to be 100mm apart that shouldn't be a problem at all. Interestingly my FSA is 4" - 101.6mm apart. SAE measurements annoy me.

By the way, I tried to add your tensionmeter to my favorites (six things listed on my profile page) but it only allows me to list my own stuff so it wasn't until I did a make that I could set it. It's set now, though.

One other point is that we change the design of the main body. That means to create some adapters "converter spacer" so you can use diferent dial gauges on one body.

For sure, I think that is a good idea to contact him and bring him here (if you can catch him), I also have other projects with higher priorities (think you have same) what doesn't mean that we can not improve his/our design.

Yeah, same here on the project priorities, the tensionmeter was a low-priority thing until I saw that you had done it all already! That threw it straight to the top of the pile.

I am trying to figure out how to contact someone on photobucket but so far have been unsuccessful. Haven't tried very hard beyond creating an account, though. I'll stay on that.

Here we go - web stalking for the win: http://www.bicycles.net.au/forums/viewtopic.php?f=34&t=90867 - going to read through that and get in touch.

Oh my god I'm such an idiot. He contacted Ric Hjertberg at Wheel Fanatyk for the original Brandt drawings. Why did I not think of that? I mean, I've referred to Ric's stuff but never thought to contact him directly for help! Amazing.

Sorry my fault, I know the thread from the austrailian forum. Back then I found it when I have also searched for the measurments respectively a printable model. But I didnt see much pictures or a upload of that, -> so I decidet to make my own design. Recently I saw the pictures by searching again for the measurements and then I have seen the printed one on photobucket and I remembered the thread but didnt found it again. I have only flown over the text in the thread, but he sayed that it should work. Yes, he contacted the guy from wheelfanatyk to get the measurements.
Back then that seems a little strange to me to contact the company which made the original ... .
But if you can get in touch with one or both of them it should help alot to get or make a cheap and easy model for printing.

Looks like cmh's internet stalking has found me. It has been a while since I worked on this project, but the majority of the info is posted on my thread here: http://bicycles.net.au/forums/viewtopic.php?f=34&t=90867

I have attached two files:

  • QV Tensiometer Parts is my take on the Jobst design. Both STL and Solidworks files are provided.
  • Tensiometer is the drawings that Jobst made.

My aim was to make a almost fully printed tensiometer. So other than the dial indicator and spring, everything else is printed. The main problem I found was over time the 'carriage' was too thin and bends due to the spring. This could be replaced with a metal carriage or one that runs on a linear bearing, but it requires more parts to be purchased which was not my intention.

Although that being said, I definitely was looking into making one that runs on a linear bearing with metal spoke anchor points so that the plastic does not wear over time. But that was once I sorted out the design of the printed version, which to this day I have not. Hopefully with you two guys and anyone else who may be reading this thread, we can further the development of the 3D printed tensiometer.

Btw Modellaner, Jobst created the drawings of his tensiometer with the intention of providing them to anyone who wanted to reproduce the design for good use(and hopefully not for individual benefit). I remember searching all over the place to find the designs, but very few mentions of Jobst's design other than on Wheelfanatyc. Ric from Wheelfanatyc was the closest person I could think of who may have access to the drawings. So an email later and I got a copy.

What CAD programs do you guys use? I'm a Solidworks user, but can convert my files over to something compatible if you'd like to make any adjustments to my design.

Hey, wellcome to this thread! I am very glad to have you here for improfment!
As descripted since I searched for a printed tensiometer I saw your design before I have designed my one.
But did not found much informations so i decidet to made "another modell" which was cheap put much precise as it could be.
In other words - it sould be a compromise. When I think back I maybe should also have to registered to the Australian Bicycle Forum to ask you about your work, that would have make things much easyer (but my english is also not well as it could be) ...
So here we are together to improve our designs thats marvelous. And yes hopefully other people will join this to.
Seems to me that you guys have much more experience (the knowledge and contacts in bicycle scene) that I dont have yet.
All I know is that Brand was a genious brain who worked on bicycle improofment. Thanks for informing me about that!
For shure, my intention to design this also is to push the international bicycle scene with a cheap but acurate tool and maybe this and other stuff will help to get new technologies ...
Anyway, I personaly use mostly Inventor but I can adapt to other programs.

Oh, wow, that's so awesome! Welcome to the thread, Quang Vuong! Thanks for sharing your design! So cool to see the different approaches to the challenge of a 3D printed tensionmeter, and cool to see collaboration from all around the world!

I am curious - how did you print the main body, specifically, the channel for the carriage? Looks like it would need a bunch of support which would be hard to remove since it's a long hole through the body.

Hey guys!
One thing that I struggle with is that the body of my design cant hold different dial gauges.
My idea to handle this is to design different spacer (with different hole patterns) that can attached to one and the same body.
So I will redesign the Tensiometer as soon as I can and now I will use the drawings from Jobst to make it small as it can be
(thanks again so much to QV for sharing his work).
One other thing came out of my mind -> To prevent the bolt that touch the spoke from deterioration you can use a little pice of PTFE-tape.
@ QuangVuong:
Maybe it also can help to make a simpel and cheap Tensiometerbody and make the slide as slippy as it can be ...

Yeah, I think it's an excellent idea to come up with a good way to deal with different mounting standards for different gauges... but another thing to address is the overall length is also different. My 12.7mm digital was longer overall than your dial gauge so I had to adjust the position of the handle on the carriage. If we designed to a standard gauge, then how do we say what gauge it is? Looks like my digital is a generic Chinese design, so I guess we could list links that we know of in different countries, but what if different manufacturers use the same body (thus the four screw mount) but different guts so the overall length is different? That becomes a challenge.

What would be REALLY cool is if we had the design parametric enough that it could be put into the customizer. Then, someone who was interested could order whatever gauge they want, then take some measurements off the gauge, and key those into the customizer and get a body and carriage that was customized to work with their gauge. THAT would be slick.

Mind you, since customizer isn't currently working, that'd be yet another problem.

Hey Modellaner, looking at that other make of your gauge, I saw he went with that same 100mm linear bearing that I found on Amazon - looking at your design for the carriage I think it wouldn't be too hard to stretch the hole for mounting the linear bearing an additional 20mm, and then it would still fit inside the existing length, yes?

I will see what I can change in new design. The "customizer" doesn't work but "openSCAD" should be a alternativ way to make the design changeable. The only problem is that I have used it only to change existing models so I dont have use it to create my own stuff. But I am shure I can handle this by train this software. Are you have any experience on that?

Had a thought - there are a bunch of designs on here that use a dial indicator, but almost all of them are for printer bed levelling. For all the designs that I looked at, they either mount to the shaft so don't need to worry about the back plate spacing, or they just model around whatever gauge they have. Doesn't look like anyone has addressed the inconsistency of back panels. So... so much for that idea. Plus, that doesn't address the variation in the overall length.

Hi guys,

I'm watching you since a few weeks, and want to chime in finally.

Right before I wanted to start the drawings of the Brandt style Tensiometer, I searched thingiverse if something like that exists already. That was at the beginning of March so, right after Modellaner posted his design :) Cool!

I had two designs in mind though,

  1. Jobst Brandt's style
  2. DT, Sapim, Icetoolz style

The 2nd is more 3D Printer friendly in my opinion. I have a printed and working model already. The friction area is smaller than the Brandt style. Wouldn't you agree with me? I'm still in testing phase but the repeatability is good right now.

A calibration device was built before the tensiometer so I can reliably work with it for my new wheels.

Regarding the Dial Indicator Gauges (German: Messuhr) there is a industry standard called DIN 878 (just search for it on google images).

I'll post more when I get further down the testing phase.

Bye

When I was researching this topic back in 2016, the two designs I found were the Brandt and DT shapes. The DT shape definitely is a more printer friendly design and is also compatible with just about any dial indicator. I could not find any designs or even dimensions of the DT design back then which is why I went ahead with the Brandt design.

Yours is looking good, so I look forward to hearing about how well it performs.

Hey guys! Sorry was not available the last days (touring with bike =D).

Great to see a new faces on this thread! And its even better if they post stuff like this!!! I deffently agree with you that the "DT-swiss-design"
is the best for printing like as QuangVuong descripted (especialy the "dialgauge mount"). I think about it when I try to design a printable tensiometer but I personally prefer the "jobst-brand-design". Nevertless I am very interested on how well it works and looking forward.
Is there a way you could upload/share your design? I defently have to try it myself :)
What kind of calibration device do you use can you tell us more about it?

The DT Swiss design might be easier to print but it works across the spoke which means that spoke readings are dependent on spoke dimensions... so in order for useful values, you would need a calibration rig. The Brandt design is good because it measures from one side of the spoke so the deflection is consistent and is essentially independent of the spoke dimensions. This is why when you see a design like the DT or the Park, you'll have an extensive tension reference sheet for different spoke thicknesses. (ref: Park TM-1 reference table ) Theoretically at least, the Brandt design could calculate tensions based on math and geometry and stuff.

For relative tensions, yes, it would still work, and honestly, for a 3D printed tensionmeter, even the Brandt design needs at least a baseline calibration if you want to know actual tension values (and I've got a plan for a test fixture that I just need to build) but for the Brandt design, you don't need such a huge reference table. I've got the FSA and it comes shipped with two columns - spokes under 2.0mm and spokes 2.0 and above. Any of the designs should give a good relative reading - ie, get all your tensions to a deflection reading of 0.53mm and you know tensions are even, even if you don't know the actual tension values.

The merit of the tension value itself is most rims have a maximum tension rating usually in the range of 95-120kgf, and if you build to your maximum tension being somewhere around there, it'll result in an overall stronger wheel because spokes are less likely to lose tension during riding, and that's where bad things happen. Having a wheel evenly tensioned to 50kgf is good, but it won't be as long-term durable as one tensioned to the rim's max tension rating.

I've stalled on this project, I do want to measure a wheel with my FSA and compare those values to what I get with the printed meter, and I want to make that calibration rig (have all the parts) but I've recently gotten a side job at a bike shop so free time keeps getting more and more slim!

@ cmh: Just came from a night ride and read your post. I didnt show up for a while because thingiverese had some issues.
Thats the facts wherefore the "Brand-Design" is best I think (thanks for enlightenment) but for sure it have also to be calibrated.
I still waiting, expectantly about results of your "calibration rig" hope you will find some free time to do this.
The scale which you tell about is available in EU but I read some reviews about it and they say its not good quality ... .
In addition I also planed to build a calibration rig with a industrial force load sensor and hope thats much acurate as possible but high quality load sesors are very expensive (especially the ones for high force with fine subdivision).
So for beginning the relativ cheap scale will do the job I think. Out of interest I orderes some cheap strain gauges from china and will see what I can do wit those (for sure they also must to be calibrated) ...

Strange, saw your comment was flagged for moderation. I saw it in the email and am at a complete loss as to why!

The perk of the Brandt design is it can be calibrated over a much smaller range. If you only build with round 2.0 spokes or round 1.8 spokes, then the DT style would be just as easy to calibrate.

Considering the price, I'm pretty sure that scale isn't the highest quality, but my initial tests with it show that it should work pretty well. A load cell is certainly the best way to do it, that's what professional calibration rigs use, but it's more complicated and expensive. My initial testing with some weights showed the crane scale did a pretty good job. Biggest problem with the crane scale is it has an auto-off function which can't be disabled and if it powers off, I'll have to remove all load before turning it back on as it needs to zero itself on power on. That could prove to be very annoying, we'll see.

I have a set of wheels I need to build for a friend, that has to happen, and I'll be able to use the FSA as well as Modellaner's design that I printed to check the values on those wheels. Not sure when that'll happen but I've had these wheels all winter and really need to get them done as he's probably wanting to ride his bike!

Not the first time on Thingiverse that happend more often (although my english is not best, I am sure that doesent depends on my spelling) ...
Anyway I remember you have telI about the "on/and load - issue" with that scale.
I found this: https://www.youtube.com/watch?v=0gdOzQdxwVQ
They only use it with a kind of screw tensioner. looks like its the same scale you have ordered!?
I am very exited to know about the differences between the FSA original and the printed one.

My guess as to your comment being flagged could only be "because thingiverese had some issues".

I just realized I never shared my tension spreadsheet, and it's relevant here. I made this way back in 2003 for the original Wheelsmith tensionmeter, then updated it when I got my first Park TM-1, and then updated it again for the FSA. It should be pretty straightforward and self-explanatory but basically you enter the reading of the gauge into the cells in the "Reading" column, and it looks up the value in the Conversion table sheet. One thing to note is I got tired of typing ".35" and so I converted the table to not need the dot, so a reading of 0.35mm is entered as "35". The sheets are protected, so you can hit return to drop to the next cell down (next spoke on the same side) or you can hit tab to jump to the next spoke on the other side of the wheel.

I've included the blank template spreadsheet as well as one of my recent builds so you can see real data.

Under "Balance Graph" there is a simple bar graph with a 0% value down the middle. The finer lines to either side represent +/- 10% with +/-- 20% on the outside of the graph. I generally build to a tension balance of under 10%, which can sometimes be quite challenging, but in the example spreadsheet you can see that built really well to under +/-4%.

Park has a spreadsheet as well, but they use a circular graph which I personally find harder to view. I've discovered that more often than not, a spoke with high tension will be adjacent to a spoke with lower tension, and that's more obvious on the linear bar graph. Plus, their spreadsheet is obviously calibrated for their tensionmeter.

This also opens just fine in LibreOffice for those of you who don't have or want MS Office. That's what I use, personally.

Wow, thanks for sharing your work! That is a nice sheet to work with and for sure will make it much compfortable to evaluate the measurments. Since I have testing my printable design I have made a sheet with a circle diagram to see approximately the differences between the spokes.
But I think your sheet is much better!

Yes think too but whatever ... Looks like the crane scale are precise enough (maybe more accurate then your other one ...).
I have ordered the same scale from a local dealer in Hamburg and hope that the delivery go fast.
Because I think I should build my own calibration rig to calibrate my Teniometer for different spoke in future.
Maybe I will have acces to calibrated and verifyed weights to see how accurate the scale work ... Anyway I think that a difference around 0,5 kg is tolerable!?

I forget to ask the question: Is the scale still accurate when its load by around 100 - 150 kg ... ?

100-150kg is right in the middle of its range, and that's where it should be the most accurate, I would think. Anyhow, a difference of a couple kg is not going to make a huge difference, so I think we should be good with this cheap option.

Okay, we can all thank my scratchy throat for some progress here. Had a bit of a cough which was keeping my wife up so I got up to clear my throat and started playing with bike stuff. Swapped in some new parts I had gotten, (yay, new toys!) and then saw one of her wheels that had a hop that needed to be addressed. Brought it to the truing stand and got the hops out, then thought - hmmm, I should check the tensions. So I grabbed both tensionmeters and got measurements off the spokes. I put the FSA values in my normal spreadsheet, but then opened another for the values for my print of Modellaner's tensionmeter. Loaded up all the values side by side, and then created graphs. I've attached a screen shot of my spreadsheet with FSA-only values and converted tensions, and attached the spreadsheet and a screenshot of that.

Notes/thoughts/observations:

  • The tire is still installed and partially inflated. Even though it's tubeless, that'll throw off the actual values, which would read higher on a bare rim. That wasn't a huge concern here as I was just looking for a range of tensions so I could make direct comparisons.
  • In the FSA values, you can easily see several high/low pairings - look at spokes 17/19, 23/25, and 20/22. Sure enough, this is the area where the hop was found. High/low pairings are wonderful as they're super easy to tune out. For the 23/25 pairing, I would take 23 down a quarter turn, and 25 up a quarter turn. This would leave the true of the rim mostly untouched and bring both spokes into a better balance.
  • I don't think the high/low pairings would be quite as obvious on a circular graph, but I'm biased.
  • Ignoring the high/low pairings, you can see the rest of the spokes are doing good within a range of +/-5-6%.
  • The FSA fits the hand nicely, it's slim and it's easy to turn the dial on the analog gauge to zero out each spoke. The Modellaner, however, due to my larger digital gauge the the longer spring I chose, is huge and a bit hard to use in the hand. Considering I've added 15-20mm to Modellaner's original design to get it to work, this isn't very surprising. Also, the zero button on the digital gauge winds up being in a place that none of my fingers on the right hand can reach, so it becomes a two-hand proposition, using the left hand to zero. The upside, however, is the digital gauge is obviously easier to read. It should also be noted that this gauge has a USB attachment which I believe allows you to zero it from another place and enter the readings directly.
  • Okay, now the fun part. Look at the graphs on the FSA vs Modellaner comparison. You could not have a more perfect correlation - the values are almost identical, just shifted due to a stronger spring that I selected for my print of the Modellaner version. Because the ratio of tension to deflection is non-linear, you can't just apply a correction factor to get a true tension reading from the Modellaner without calibrating it, but the key part here is it works absolutely perfectly.
  • This means you guys have duplicated a gauge that was something like $250USD with about $50 in parts. Once I get my new printer (any day now) I want to try to print QuangVuong's gauge and see how that compares. I'm expecting it'll also work quite well.
  • With better choices for the spring selection (mine is too long and too firm) I think the thing would be far easier to use and closer to the FSA in deflection numbers.

This is really cool, I thought it would be good, I thought it would be close, but this is way better than I expected.

Wow okay that are great news :D ! Agree with you that a bar graph like in your sheet is best for that. As I have made the circular graph, I only want to see the difference between the spokes and did not think about a bar graph that compare that much better (especialy like your worksheet) that's exactly what this thing is about: tension balance. So in the name of everyone thanks again for sharing your work!
Hope we can find a way to calibrate the printed one's as soon as possible (maybe not only with different spokes rather with different springs ...) . I am also exited to see Quang Vuongs - design in comparison. If it works nearly identicaly than the FSA its best for 3d printing (because it is cheap) I think (@Vuong: hope so much that you share your design as "official" some time).

No progress on this since May, but I haven't forgotten about it. We're losing light here and riding is happening more and more indoors and that's usually when I find myself focused more on 3D printing stuff - I still need to make the calibration jig and I had a couple ideas about mounting the dial indicator which might prove to be a bit more universal. Hoping I can get to that at some point soon. Plus, I still want to try printing Quang Vuong's design from the Brandt drawings. Hopefully this winter we'll make half the progress we made last winter!

Same here the autumn is there so that means winter is near for more sparetime for 3d printing and other stuff. As I just read your comment I was laughting because some days ago I was in my workarea, grabbed out the crane scale and think about this project especialy about the calibrationstand and how to realize it ... . A more universal mounting for the dial gauge plus Quang Vuong's design of the slide will be very helpfull to make it easy replicable!
So yes I hope we can push this further.

I was starting to look at the files that we got from Quang, and I'm having a bit of a time opening the DXF files. Best I've managed so far is inserting them into LibreOffice and then adjusting the brightness all the way down so the yellow text shows up as black and can be printed. I can import them into Fusion 360 but it just gets the geometry, none of the dimensions. AutoDesk has a number of online viewers, but I can't get them to open the DXF files, either. Even with that hack of importing into LibreOffice, springcup.dxf appears to be empty and capstrt.dxf looks to be cropped. Did you have any better luck opening the DXF files, and can you maybe convert them to PDF? I'm attaching the PDF I did manage to create.

Normally you can use for example "Autodesk DWG TrueView" or similar to view the DXF-files. I think you want to view the files to see the original measurments of the design!?

I tried to download DWG TrueView but discovered it's windows only! I've got no Windows systems anymore, I'm all mac. I could load up a VM that's an ugly solution. I can see most of the original measurements from my PDF but it doesn't look like I'm getting everything when I import those DXF files into LibreOffice or Fusion360, so I was just curious to see what the drawings are supposed to look like.

Ah okay thats the thing with the different operating system platforms I know ... :D So then I will see if I can convert them to .pdf and load them up here. Do you have a imagination of your prefered design from the coloration of the sheets? Because in original they comes with dark background and yellow drawings and scripture. But I think I can convert them to a different design if you need one.

Whatever works - if you can convert them to a PDF which can be opened, I'm not worried about the exact colors. I just want to see if I'm missing something from what I can see when I import. Thank you!

Okay after trying to convert from DWG-Viewer and other programs I used to make screenshots out of the viewer import them in openoffice and convert them to .pdf. So thats the primitive way but it works well I think. Hope you can work with that!?

Holy crap, I just looked at the date on the prints - 05 Nov 1975. Those tensionmeter designs turned 43 today!!

Beautiful, thank you!!

I am glad to read that :D Thanks to jobst brand and quang vuong for pre work!

Regarding the spring, I see the original Brandt drawings spec a Lee LC-032E-15 compression spring which is 0.360" OD, 0.032" wire diameter, 2" free length, and 4 lb/in rate, which matches the music wire on the link. I have found a similar stainless steel spring on Amazon but that's only a 3.3 in/lb rate. Looks like this music wire spring is exactly what we're looking for. Going to take another look at the digital gauge and see if those spring dimensions will work, otherwise I'll look for something with a comparable range and rate.

Okay, latest update, I was thinking about our mounting issues with the variance in patterns in the back of the gauge, and I think this addresses that. Leave the bolts off the back of the gauge, and clamp to the shaft as is done with that adjustable arms often used with this style of gauge. I'm assuming that most gauges have a 9.5mm bore diameter similar to the one I've got, and if that's the case, this design will mount them all without too much trouble. I've sized the hole for 58mm bore which should fit all the sizes we saw if I remember the numbers right.

I also liked Quang's approach of swapping out the bearings for simple cylinders. Would be easy enough to have two versions, one with the cylinders and one with proper bearings for folks who wanted to be fancier about it.

Thoughts?

Awesome!! Looks like a nice prototype. Thats the approach we need to improove the design. Hope you can share the new version soon!?
Change the bearing slide to a simple cylinder will be not as complicaded as it seems. It only must be strong enough so it can't bend.
I see you modified the plate on the bottom and remove the tunnel for better printing!? Maybe it won't need a guide to be in place.
As you described: "the pressure point it is on the same axle as the dial gauge so it should be stabilized by it self" and the other point at this is that without a plastic guide there should be less friction.
So I will look forwoard and see what improvement you have next.

Check this out - WheresWaldo made a spinoff of the DT Swiss tension meter

Looks like a nice, clean design, too, simpler to print for sure, but those designs need to be calibrated for each spoke type.

I had been thinking of doing a DT knockoff once we got a little farther on this project, but nice to see that someone's already done it!

Spoke Tension Meter

Still very much a work in progress - about the only usable part right now is the gauge clamp. Still haven't figured out the rest of it, how to address the flex in the back piece, etc... Either way, here's a link that should get you to the online version of my work so far: https://a360.co/2PcaNK6

Of course all of this is because it's easier to do stuff in CAD than to build the tension calibration setup that really needs to happen.

Latest update - I have been doing a bit of research on linear dial indicators, and the one thing that seems consistent among them is the 9.5mm (technically 3/8") shaft for clamping. The other thing which is consistent is the back lug - but, as we've already found, not the back plate mounting. So, with that in mind, I think clamping on the shaft is going to be the best approach. The original Brandt design specified a Starrett dial gauge by part number, so he could work with the back mount size of that gauge.

As it is, however, I think that with variations in the length of the center rod and overall length of the gauge, the design is going to have to be unique to each gauge. Don't know as we'll be able to get around that, and that kinda sucks, so right now I'm still working off that digital gauge since it seems to be a commonly available part under different brand names.

For the overall project, I'm now under a little bit of a time constraint since a coworker at the bike shop is going to build his first set of wheels, and it'd be cool to have a functioning tensionmeter (which has been calibrated in the fixture I still need to build) for him to use to check his work. Hopefully I've got a bunch to report in the near future!

Oh wow looks like it goes on and on with this project. I was not availabe the last days because touring in germany. But nice to see that you made progress in this. I also took a look on the DT-design from "WheresWaldo" that is also greatfull.
Maybe I also will find a little sparetime to build my own low cost calibration device... .

  • So yet it looks like you finished the new brandt-design (as I saw on the Fusion 360 files)!?
  • Do you think that it is ready for printing?
  • Are you making progress with the calibration fixture?
  • Hope you finished it until your deadline!?

No calibration device yet but that's the only bad news we've got. As you've no doubt seen from the Fusion 360 link, it's ready and it's looking pretty good! Have what should be the last print running right now to address some minor issues I found with the bar and handle, but nothing that stopped the version in the photo from working. Thanks to the spring spec in the drawings, I was able to get the original spring, and checking between my FSA and the prototype, but have about 4lbs of force (2kg, ish) at the zero point. I'm going to try to get a thing posted today. The fit in the hand is pretty good, it's still a bit bigger than the FSA but a large part of that comes from the gauge being a 12.7mm throw - the FSA gauge is only a 5mm throw, so we have to deal with almost 8mm of wasted space. One thing I'm considering for a future revision is modifying the digital gauge to chop off all but 5-6mm, which will make the overall design much more compact.

So now I really have to get serious about building my calibration fixture! That should be the next thing.

Posted my version: https://www.thingiverse.com/thing:3220573

I've been trying to figure out how we would make this design as gauge-independent as possible, but I don't think it'll be possible. Even with 12.7mm gauges, there doesn't seem to be any standard to the length of the clamp area, or the diameter of the gauge body, or the overall length - all of those would affect the geometry of the body and the bar.

Bicycle digital spoke tension meter based on Jobst Brandt design
by cmh

I am printing yours now. I am still fine tuning my printer profile so I decided to use ABS. So far it is just the main body. Downloading and importing into Fusion 360 is just tedious, but that is Autodesk's fault. This is being printed AS-IS, but my slicer settings will not print the threads fine enough, so I am going to rely on a interference fit for the screws.

Looking good so far! That's surprising about importing into Fusion - you'd think that sharing a cloud-based item would make it easier to import it, but to be honest I haven't done anything with sharing designs in Fusion 360 (only just realized that it's cloud based so the stuff is available online) so I might not be sharing it the best way.

I think the interference fit should be fine - if it's a little too tight or hard to start, I find opening the hole a little with a file can go a long way to get it to start, and then it should thread from there.

One other idea I had after looking at @Modellaner's comments on the problems he ran into with my design was maybe we change the whole approach - instead of a big triangle shape that holds the gauge, why not just have the bottom of it that has the finger holes, the clamp for the gauge, and the spoke measuring area - basically just the bottom of the current design? That might make it possible to make the design a bit more gauge agnostic, and keeping the bar closer to the centerline of the spring and the gauge might reduce the amount of flex. Don't know how much that'll help - as it is right now, I've only got a 0.5mm wall thickness behind the gauge and another 0.5mm from there to the bar, so not a whole bunch of wasted space to be reclaimed.

Printed and built.

There are some things I like and other things I don't. Unfortunately the things I don't like would keep me from using it on a wheel build. It is much easier to hold than the others on Thingiverse. and seems more stable in the readings. I like the minimal parts count.

I could not get a spring that sit on the Spring Seat that also fit the recess of the Handle.

I also might have designed the slide differently, maybe with the spoke rest as a separate piece instead of the handle as a separate piece. In addition, there is a tendency on the one I built for the slide to start moving askew from the main body, not parallel to it. Not an issue with @Modellaner's meter because of the linear slide he uses, maybe there need to be a guide of some kind or another on the rear of the main body to keep this from happening.

For another design idea, I am going to see if I can't pick up a tire tread depth gauge, and design something around that. I think it could really minimize the entire meter and they are all basically the same size as they are only the display portion of a digital caliper.

The dependency on one particular size of spring is frustrating, I definitely don't like that.

I thought about the anvil as the separate piece vs. the handle, that's still something that could be done. I still don't think the linear slide or some sort of guide is necessary, as that wasn't present in the original Brandt drawings, it's not present in my FSA, and doesn't appear to be present in Wheel Fanatyk's current digital offering. The key is the spring seat that fits into the handle and acts as the guide on that end. With that in place on mine, the action is easily as smooth as my FSA which is mostly machined aluminum parts on a carbon fiber body.

Now your idea about a digital tire tread depth gauge, that's a brilliant idea - can't wait to see what you come up with! Looking at a couple I see they can be had for about $10 USD, so if you can mount it well I think it'd be a great design.

A good calibration unit is what we need. Then there is no dependency on any particular spring or dial indicator, or Tensiometer design either.

For one to be usable, it has to be cheap to build, require only hand tools, be repeatable in measurements. My guess is that you will need to realistically pick any two of those.

At least that is my goal in the other thread, but I have already failed the requirement of only using hand tools.

Comments deleted.

Agreed on the calibration unit - have a design in my head that might fulfill the requirements and need to put it together. Assembly shouldn't take too long, but some other things are taking priority at the moment so my progress might drop off for the next couple weeks.

My first version of the Calibration Rig is done. It works, cost about $80 USD to build, only needs a hand drill and a few hex wrenches to build.

I found an interesting thing about springs and tensiometer design. The DT Swiss style requires a much smaller spring to function than the Brandt/FSA style. The two Brandt closes I built, I had to take apart to replace the spring with a stiffer one. The DT Swiss clone worked with the little spring I used.

My calibration rig goes to 1476 newtons and when I compared it to my Park TM-1, the Park Tools read higher tension than actual by one mark. turning the spring seat 3 whole turns got it calibrated perfectly at 1100 newtons. The Park Tool Chart I have is the old version and for some reason Park Tools still uses kg.f instead of newtons and their assumption for calculating newtons is off. Park uses kg.f X 10 when the actual formula is kg.f X 9,80665, insignificant at low tensions, but a wider error at high tensions.

$80 for a calibration rig that uses a load cell is seriously impressive!

I'm curious, why did you find you needed to increase the spring in the Brandt style designs? One of the design features is that it uses a particularly light spring and only creates small deflections. The spring I used matches the rate in my FSA, and both of them only show between 0.3 and 0.4mm on a 1.8mm spoke with 100kgf tension in it. Were you trying to reach a certain deflection?

Really surprising that Park got the conversion wrong, like you said, it's not a huge difference but it's an added bit of inaccuracy. Maybe they figured that was close enough for the less precise nature of the TM-1.

Regarding kgf vs. newtons, kgf has been more commonly used by rim manufacturers in terms of specifying target tension. It's what my original Wheelsmith used, and it's also used by the FSA. Looking at the meters available on the Wheel Fanatyk store, both the P&K Lie and the Wheel Fanatyk models come with calibration charts in kgf and newtons, so it looks like newtons are gaining some traction. Makes sense, as they are the actual unit of force in SI units.

Looking forward to seeing the setup!

Already pictured in the other thread. To be clear, I built the cheap version first and it uses a Crane Scale (of course the scale has a load cell in it). With regard to measurement scales kg.f is deprecated as a measure of force, it basically had no meaning, newtons is the correct scale and at the very least DT Swiss' current documentation lists newtons in their charts. My old TM-1 Chart (the all blue highlighted one, not the new multi-colored version) specifically says 1 newton = kg.f x 10, which for most purposes is probably close enough. I just checked, the newest version located here uses the same formula.

Regarding spring selection, I think I was using an undersized spring to begin with. I used what I had, since I didn't want to invest more money in parts. My priority was the calibration rig.

Pricing, I am pretty sure the load cell one could be built for about $125 not $80. My version will cost more because I used 3030 series Aluminum T-Slot even though I now realize it was unnecessary.

So your choices are, go ghetto and use a hanging scale, or go luxury and use a bare load cell and digital indicator. My guess is that the bare load cell version, which uses a bigger load cell, should be more accurate. But I am also sure it really won't matter in real life.

The tire depth gauge version is going to wait a bit. Spent all my discretionary cash on Chinese parts for the calibration rig. Got a load cell, digital readout, 3030 extrusion and fittings all coming from China (see other thread). I still think it will be easy to do, but I need at least one tire depth gauge in hand so that I can design against it.

Well, no rush. Modellaner will tell you we were all types of busy on this project last year and then spring came along and it was time to ride outside and nothing happened until recently.

Yeah! That's the way it is. Exept the fact that I ride long distance tracks also in winter ;D So at the moment I don't have much time for this project because I am mostly on the road (not only with my bike ...) But it's very nice to see that others make the progress on these projects ! :)