Someone tensile tested 2 cheap GT2 belts, the springiness is significant: https://www.reddit.com/r/3Dprinting/comments/61j9et/i_tested_some_gt2_belts_to_failure_on_a_tensile/
@Sven_Eric_Nielsen I’m 100% with you. But I don’t know the properties of the 2GT belt, because it isn’t branded. And until now, I don’t know the specs of the later used GT3 belt either (I have asked http://gates.com directly, also the vendor I bought it from, still no answer) So, noone knows for real. I know about a test a german http://reprap.org user did, very similar to the one @ChPech linked. He tested a noname-6mm-2GT vs an MXL-9mm. Not very fair, I know, but it did show the overall “stretchiness” of the 2GT belt, too. Graphic attached. What I want to say: It is striking, how much the nonbranded 2GT belt stretches, constantly, until it snaps, right from the start.
And if anyone can use google better than me, pls link me the datasheet to the “Gates Powergrip GT3 2M-6mm”. Thanks.
Amazing to me how many people have apparently tension-tested garbage off-spec belts without bothering to compare a genuine belt. The hell is the point? You’re not going to get any consistency or repeatability with trashy belts. Two different off-spec belts will test different, and there’s no way to tell what you’re getting when you order. The only thing you’re really learning is that cheap garbage has inconsistent and poor performance, which should be blisteringly obvious.
Why doesn’t the manufacturer provide those? Like Gates or similar?
@Rene_Jurack they’ll provide data on request to people with a legitimate need for data. Like if you want to select a belt for 100,000 doodads, they’ll help with that. Don’t think they have too much interest in giving out detailed performance specs to the whole internet. It’s not the sort of product where providing lots of public data is going to get you much competitive advantage.
Check this though:
This is the sort of data authorized resellers can get.
Based on the published belt specs, this is how genuine fiberglass core GT2 should approximately look overlaid on the Reddit tensile test chart. (6cm long test sample.)
Should be pretty clear that one knockoff belt tested was fiberglass but low quality (premature failure) and the other knockoff had a polymer tensile member (polyester or nylon) which is completely unsuitable for 3D printers. [EDIT: details in Reddit thread said the junky belt had the elastomer separating from the tensile yarn (junk!) so the tensile member wasn’t carrying load effectively.]
missing/deleted image from Google+
@Rene_Jurack I did some math on theoretical belt stretch vs motor error, and with the DICE approximate hardware specs, I’m calculating that belt stretch with genuine GT2 is about 4-5x bigger than the stepper’s rotor overshoot at the same drivetrain force. (Genuine GT3 should be slightly stiffer than GT2.) So switching to 0.9 degree steppers (or bigger steppers, or smaller pulleys) would be a marginal upgrade. Might still be worthwhile, but another belt upgrade (eg to 9mm) should be more effective.
Of course, the challenge with these calcs for CoreXY is that you can have such a wide range of moving masses and belt lengths between the motor and load that it’s really difficult to get good inputs. For example, one direction of motion near a corner with a motor might have 100mm free belt length (in which case belt stretch is very small) while the opposite direction of motion at the same position might have 800mm free belt length.
https://www.gates.com/~/media/files/gates/industrial/power-transmission/manuals/light-power-and-precision-manual.pdf
Page 59 for GT3 specs. GT3 is about 9% stiffer than GT2.
Can anyone provide a good US source for small quantity continuous length Gates GT3. It is very expensive stuff and the Amazon 5 meter offeringsv would result in tossing a lot in the rubbish bin for my delta 3dp.
I bet you have a us-shop for timing belts. Germany has three I could find within 10sec google-search…
I usually check SDP-SI stock for belts first. They don’t always have stock of everything in their catalog though.
Thanks for the reference. But 12 weeks backorder!! Jeez
I took your advice and finally managed to swap my belts with GT3. I increrased the acceleration to 3000mm/s², movement speed to 300mm/s and printing at 200mm/s. The ringing was minimal so I might try more acceleration, although it is an e3d BigBox with a direct drive extruder, so the moving mass is huge.
The downside is that my volcano hotend can’t keep up with the extrusion rate, depending on material 30-40 mm³/s is the limit then the poor 40W heater cartridge is just giving up. The second downside is it is very loud despite using TMC2660 drivers.
@ChPech You are the first of all testers and you confirm my observation. Thx for sharing.