2

I have a 3D printer that is going crazy with x-axis shift, and I need a tension gauge belt to measure the tension.

I've never used one before, and looking online, I can't tell which one would be the right fit.

Any ideas? What things should I look for?

Trish
  • 20,169
  • 10
  • 43
  • 92
Bashar Abdullah
  • 123
  • 4
  • What kind of printer is it? Most Cartesian printers that experience X-axis shift are having problems with fatigued electrical wiring to the X stepper or X endstop. – Ryan Carlyle Apr 12 '16 at 14:35
  • @RyanCarlyle It's Cubicon. Not popular brand in West. Korean made. Supposed to be good. But soon after getting it, this started happening. Part of their troubleshooting suggested checking the belt – Bashar Abdullah Apr 12 '16 at 20:17
  • Looks like a cartesian XY gantry printer, from a quick googling. If you can pluck the belt like Tom mentioned and get a low note, but still get X axis shifting, my next guess would be bad X motor or X endstop wiring. – Ryan Carlyle Apr 13 '16 at 02:04

3 Answers3

6

It's extremely unlikely that belt tension is actually your problem. I've never heard of anyone using a gauge to measure their belt tension. Typically you just pull your belt tight by hand so that it produces a low note when plucked. It's far more likely that you're experiencing shifts due to too high or too low stepper current.

Unless your belt is so loose that it easily skips over the pulley (which should be obvious without using a gauge) or so tight that it completely binds up (it would be impossible to get it that tight without some kind of superhuman force) it's definitely not the problem.

Tom van der Zanden
  • 14,588
  • 2
  • 33
  • 60
  • 2
    I've read of people using smartphone apps that calculate the belt tension by measuring the frequency of the plucked belt. If you (the TO) are completely lost and need any kind of confirmation this might be a way to get it easily. – kamuro Apr 12 '16 at 10:00
  • 1
    I used a spring scale (meant for fishing). But it's really not critical. Just make sure it's not skipping teeth on the pulley and you should be fine. – TextGeek Apr 12 '16 at 17:34
  • 1
    I tensioned the belts of my delta printer equally tight solely based on their note when plucked. If you have an ear for music, that should be fairly straightfoward. – Tormod Haugene Apr 12 '16 at 18:42
  • @kamuro thanks, will do if needed. Although now it seems not the issue. – Bashar Abdullah Apr 12 '16 at 20:18
  • @TormodHaugene alright. Although now it seems unlikely this is the issue. – Bashar Abdullah Apr 12 '16 at 20:19
  • @Tom van der Zanden Any tips how to know if this is stepper current problem? I'll most likely end up opening separate thread for that. But it would be good to have more info. – Bashar Abdullah Apr 12 '16 at 20:24
  • 1
    @BasharAbdullah your issue is one of possible signs of current problem :) #1. Check resistor on your SS (right next to black central chip. #2. There should be R200 or R100 so R=0.2 or 0.1 #3. Read current (I) value from motor label #4. V = I * R * 8 #5. this is max voltage which should be set on SS #6. set it and check if your printer works ok #7. if not - reduce it 10..15% and chceck if this helps. you can repeat #7. – darth pixel Apr 13 '16 at 08:54
  • @darthpixel thanks. I'm new to 3D printing really, so will try to tinker and see where I get :) – Bashar Abdullah Apr 14 '16 at 06:51
3

You may not need to adjust the tension, but you may nevertheless want to measure the belt tension, at least to ensure it's not too tight: the shaft on stepper motor has limitations.

In that case, no need to buy a tool, you can use an audio app showing the frequency spectrum and then you pluck the belt. The lowest frequency peak can be related to the tension as explained here:

https://benchtopmachineshop.blogspot.com/2019/04/printer-belt-tension.html

For GT2 belts and 27 N tension, assuming the distance between belt and motor face is 10 mm, the relation is

$$f [\text{Hz}] = \frac{28531 \frac{ \text mm}{\text s}}{length [\text{mm}]}$$

Be aware that by plucking the belt you will excite the higher harmonics more than the base harmonic, therefore identifying the lowest frequency may take some time and multiple tries. You should display the frequency spectrum as "waterfall". The app "Spectroid" for Android has it.

If the distance between middle of the belt width and motor face is twice the value assumed (for example, 20 mm instead of 10 mm), the belt tension should be halved (based on an inverse relationship).

FarO
  • 3,906
  • 13
  • 31
  • 1
    heh, tuning your belts by sound? I always was going for a *nice* sound, not a specific frequency, but this is gold! – Trish Oct 12 '21 at 21:27
1

People are using this Clothpin springs repurposed into a makeshift belt tensioner to tighten the belt but I would totally agree with @Tom - there is no need to use such devices or gauges and your problem lies in stepsticks current.

It would be also problem with endstops so PCB doesn;t know it's time to stop :)

Trish
  • 20,169
  • 10
  • 43
  • 92
darth pixel
  • 3,438
  • 1
  • 11
  • 20
  • Can you explain how this belt tensioner (if that is a correct name for it) works? – kamuro Apr 12 '16 at 10:01
  • 3
    @kamuro it's just a spring that you put on the belt. It isn't really a tensioner (you still have to pull the belt tight somehow) but it helps keep the tension consistent. – Tom van der Zanden Apr 12 '16 at 10:03
  • 2
    In the meantime I found the idea by image searching 'belt tensioner', too: http://www.iheartrobotics.com/2009/11/timing-belt-tensioner.html The search also turned up lots of other interesting tensioning designs used in various other (industrial) sectors, too. Sweet inspiration ;) – kamuro Apr 12 '16 at 10:05
  • 1
    On another note, I would assume that the current of the stepsticks also is a less important effect. I would take a look at the axis motion friction first. Parallelity of rods first, if the strength of the effect is varying depending on the position. – kamuro Apr 12 '16 at 10:09
  • 1
    Well - depending on drive train design - this effect can also apear because of low wrap angle. Which means that timing belt surrounds a pulley not enough. – darth pixel Apr 12 '16 at 11:32
  • 1
    @kamuro as TvdZ said - it's not real tensioner it's more like laundry clip but it works fine so why not to give it a try :) But be careful it cost even $2 (as a "tensioner") but if you buy 10 laundry clips you will get 10 "tensioners" for half od dollar. – darth pixel Apr 12 '16 at 11:38
  • I didn't want to diminish the idea of using a belt tensioner at all, sorry if it came across like that. – kamuro Apr 12 '16 at 11:41
  • @kamuro not at all :) – darth pixel Apr 12 '16 at 11:48
  • 1
    I would advise AGAINST using that kind of belt tensioner. Springs anywhere in your drivetrain add ringing at corners. The last thing you want in a precision motion control device is stretchy belts. – Ryan Carlyle Apr 12 '16 at 14:35
  • @RyanCarlyle Are you sure about that? It seems these spring-type tensioners are used in "professional" printers (Ultimaker comes to mind) and even some mass-market 2D printers. In any case, this problem is likely not about belt tension. – Tom van der Zanden Apr 12 '16 at 14:44
  • @RyanCarlyle don't you think these tensioners are just compensate sudden de/acceleration. Belt itself is not stretchy, its dimension change during acceleration only then it gets back to normal so counting whole amount of the movement dimension doesn't change a bit. Timing belt itself is doing the same by its nature isn't it? – darth pixel Apr 12 '16 at 16:45
  • 3
    Yes, I'm sure. A spring-tensioner is better than a loose belt, which is why some Ultimakers have them (particularly since getting even belt tension on a UM gantry is hard) but a rigid tensioner (adjust and lock) is superior to a spring tensioner. We use fiberglass-core and aramid-core timing belts in 3D printers for a reason, because they're as stiff as possible in tension. Spring tensioners makes the belt more stretchy. Specifically, when you turn a corner at positive velocity (jerk>0) any springiness in the drivetrain causes increased overshoot and therefore ringing. – Ryan Carlyle Apr 12 '16 at 17:01
  • 1
    Should add here that if the clothes-pin spring tensioner is nearly fully extended, it will be fairly stiff, because the geometry creates a force multiplier to make the spring apply a lot more tension. (EG see http://physicsbuzz.physicscentral.com/2015/04/physics-life-hack-number-3-getting-your.html) But if the spring tensioner is taking up a lot of slack and is not extended, it will add springiness to the belt. – Ryan Carlyle Apr 12 '16 at 17:25