I was looking for 125Hz, which I picked because that is what you used 
Once I found the “tooth” that got me close, I then used torque on the bracket screw to dial it in. It wasn’t easy, took 30 minutes of trial and error until I got it right. The shorter Y belts (I have an XL) were much easier than the longer X belt. Those with the XXL will probably need to budget more time.
Yeah, it’s very fiddly the first few times, once you’ve done it a few times it gets faster. Try video recording yourself explaining what you’re doing until you get that clean and then you’ll find you’ve learned to do the belts without thinking about it 
I would be interested to hear about what kind of calibration factors ($100/$101) you folks end up using after using this belt tensioning method with a target frequency of ~120Hz.
I had recently reset my machine to default GRBL params, which among other things resets $100 to 40 steps/mm, so I went to recalibrate it only to find that…it was pretty much spot on (I cut a 30x30mm piece using a finishing pass, and got 30.00mm on Y and 29.96mm on X). It could be a coincidence, but…
I think my X (steel-core) belt is telling me that it has had enough, it won’t twang for me no more.
Either 125Hz was a tad too much for SO3’s X axis, or most likely it just wore out, it’s been a good while since I changed it.
So, it’s an ex-belt now?
As for Y that happened, well, Gates has quite a long discussion of the relative merits of the different core materials for the belts, glass fibre and aramid fibre both score very well on the “keeps bending” scale whilst steel can survive excess tensions better than either of the flexier fibres.
However, the max rated tensions for the OEM Gates belts is in the 800 Newtons range so I suspect that’s not what broke it.
I think the steel cored belts are best just considered a consumable like cutters and an item you regularly check for consistent tension tone, keep spares of and be prepared to replace. I think the Carbide and AliExpress steel belts are cheap enough compared to the OEM gates fibreglass belts to be treated as a consumable too.
And no, I’m not at all sorry for the puns 
Also,
I’m continuing to review the evidence on belt tensions and I think that table needs revising soon to stop not far above 100 Newtons for the sake of the steppers.
Yup. I’m still on the first roll I got on Ali two years ago, and I know what I’ll be doing tomorrow evening.
@Julien
Just curious,
in another thread, it is mentioned about the steel reinforced belts -vs- the bend caused by the radius of the idler pulleys and stepper sprocket. I think I read that this tight of a bend is not “Good” for the steel. … What about the folding the belt over to use the belt keeper/tensioner.?
Have there been any belt fails that may be caused by the steel wire breaking at the fold in the clip and causing what would look like the bad belt pictured above?
It seems like that would be the likely area for the belt to fail from over tension…?
The reinforcing members are bonded into the neoprene over the full length of the belt, otherwise cutting the belt to length would result in a bad outcome…
So far as I have been able to determine from the manufacturers info, the issue with small radii for steel and fibreglass reinforcing members is the bend radius coupled with repeated flexing causing the reinforcements to fail at the flexing locations.
The belt clips are way below the bend radius any of the belts are meant to deal with when moving, but they’re stationary, so there is no repeated flexing to cause a wear failure. Whether a compression grip tensioner like the ones Neil Ferreri posted;
are a better way of gripping the belt and it’s internal reinforcement members I would like to know.
Maybe the tugging and loss of tension during normal operations? Constantly being “flexed” to some degree…
https://www.mpja.com/GT2-COG-Belt-Aluminum-Clamp/productinfo/33400+HD/
Yeah,
I’ve been thinking about belt clamping and looking at a few options, I’d like to see some failed belts in close up detail to see how they’re actually failing and where.
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.
Thanks to Julien for updating the belt tension table in the first post of this thread.
Back in 2020 I only included tension for the 9mm belts as there were no (unmodified) machines running a 15mm width belt. Unfortunately the 15mm belt in addition to being wider, is heavier per unit length, which means it has a lower frequency than a 9mm belt for the same static tension.
This means that using the same frequencies as originally posted for 9mm belts on a machine with 15mm belts will likely overtension the belts quite a lot.
Apologies if this has caused anyone any issues but I only recently looked at the calculations again and realised how far out the table would be for 15mm belts.
I’ve seen no evidence yet that increasing belt tensions beyond the minimum required to match stepper forces improves performance, whilst we know that there are downsides, so here’s the updated table which includes the 15mm belts.
Select either 9mm or 15mm belt
Then whether you are measuring over 250mm or 500mm
Then pick a target tension, read across to the target frequency
e.g XXL machine measuring over 500mm with 15mm belts
Choose 89N → target frequency = 58Hz
I use about 95N static tension on my SO3 XXL and it is fine, calibrates well etc.
I have put the 150N in a warning red as this is likely more than enough to upset the stepper motors on the 15mm belt machines with their wider pulley providing greater leverage on the shaft.
Hi Liam,
I have the Shapeoko Pro XXL and have downloaded a frequency measurement app for my phone.
If visualized as a ‘stringed instrument’, to which frequency shall I ‘tune’ the X axis belt and at which frequency shall I ‘tune’ the Y axis belt?
Thank you in advance,
Pablo
OK,
So assuming you have the 15mm belts on your Pro and you’re measuring over 500mm width between supports that’s 58Hz to 67Hz.
According to some smart people that’s a B1 or C2 note, does that help?
Yes It does. Thank you Liam.
Then, how do you ‘hit’ the note?
Not sure I properly understand the question, if you mean how does one pluck the belt to get the tone, here’s how I do it
If you mean how to actually tune the belt, I leave a turn or so on the M5 bolt on the belt tensioner to allow for some fine tuning, if there’s not enough there then I loosen it off and move the belt a tooth in or out.
Thanks for the updated table with the 15mm belt values, it’s a big sanity saver.
I will note that when I tension the belts, I find it easiest to do the initial connection with the machine at the center of all axes, then I’ll slowly/gently move it to one end, then power up and have someone pull on the gantry/carriage so as to take up the tension which makes it a bit easier to get the belt tightened up and secured, then power down, and repeat the process for the opposite end of the belt.
