I think most people where actually overtensioning the steel belts because they stretched so much less than fiberglass. Forum members tested this and it was a big number, something like 50% difference.
Something interesting to think about is that steel belt failures are pretty easy to see with a quick inspection. Fiberglass ones seem to go in stages.
It will be interesting to see how the pro performs with its 15mm belts.
I am looking for Nema 23 motors with a longer shaft so I can run 15mm or possibly 20mm belts but instead of the stepper nose bearing and shaft taking all the stress. I want to add a bearing block so the shaft is fully supported on both sides of the pulley. It would allow for wider belts, stronger steppers, and prevent snapped shafts while adding even more rigidity and possibly a longer motor life span.
The design pattern for this is to use a separate shaft with two bearings for the pulley and then use a shaft coupler to the stepper. There will be little choice in steppers with long shafts because that’s not how the mfr expects them to be used.
If you want to go up in radial load spec then the first thing is to go to 8mm or better shaft size.
Just curious, which method did you (@Vince.Fab) use to tension your belts? There are a few methods on the forum, so I wanted to see which you use.
Also, you mention a 6-8 months lifespan for your belts, how many hours is that? And do you notice more wear/breakage around the x-axis pulley in home position? Is your breakage always in the same location?
Sorry for all the questions Vince, I love your contributions to the forum, you (and others here) do awe-inspiring work!!
I base everything off feel and real world accuracy. Although I’m not against the higher tech techniques, imo the simpler the better.
Steel cores go 0-100 really quick, and I’ve only broken one set actually. Location seemed to be random. Given how much stiffer than fiberglass and the ability to be accurate without micro adjustments, its a no brainer. No ideas on total hours but my current set has over a year on them, plus long lengths of time fully tensioned.
Lots of talent and brain power on this forum for sure! Not just a bunch of button pushers
I’ve got over a year on my steel belts, and have never calibrated GRBL or adjusted belts. I have had to do some minor tweaks over time in fusion to account for what I’d assume is minimal stretching, has worked fine for what I do. I design everything to size, then use compensation to tune fitment. Most apparent on small adaptive and boring operations. Plan is to adjust and calibrate everything this week while swapping some things on the machine.
I’ve got about 9 months, including some abusive testing on my kevlar cored belts which seem to be a less brittle alternative to the steel core whilst doing less stretching than the fibreglass varieties. ( from reprap.nl )
@LiamN I received 5 meters of the Kevlar from reprap as well (back in March), but I am hesitant to put that on because of the stretching @mikep experienced with his, and that occurrence seems to be typical for aramid/kevlar reinforced belting:
…However, it’s important to note that while Kevlar has excellent dimensional stability in dry and stable environments, wet environments, or those with fluctuating humidity levels, can be detrimental to Kevlar-reinforced belts. This is because Kevlar cords (and cords made of other aramid materials) experience a change in length as they absorb and release moisture. This length instability can affect the belt’s performance — altering the engagement between belt and sprocket teeth, or affecting the tension of a V-belt, for example.
Although Kevlar is suitable for a wide range of temperatures, it has a negative thermal coefficient of expansion, meaning that it shrinks as the temperature rises and lengthens as it cools. Since this behavior is the opposite of most other machine components (especially those made of metal) — which expand as temperature increases and contract as temperature decreases — Kevlar-reinforced belts may not be suitable in environments with significant temperature fluctuations.
My SO3 is in a location that has temp swings of ~10-15 degF over the course of the year (August max to February min), and humidity fluctuations of ~35% relative humidity in the winter to ~60% in the summer…
It’s swings and roundabouts, mine is in a shed which regularly changes 15C every day, I periodically re-check the tension with the phone app and re-check the steps/mm when I have to re-set the tension.
I’ve only had to re-tension and recalibrate after some particularly violent abuse involving a fractured cutter in Aluminium making the whole machine shake.
EDIT - Swings and Roundabouts = Snakes and Ladders / Benefits and Problems for the non English
I think I also mentioned that as an alternative to fiberglass and steel I really liked the kevlar, and have moved all belts on my machine to kevlar. The fiberglass stretches a lot more, and the steel has fatigue problems I don’t like. Kevlar is better. They all have tradeoffs.
One purely better option over the generic fiberglass belts is to get actual Gates-branded belts — quality/consistency seems much better, as does lifespan — I’m still on the set I put on my Y-axis when I did a 9mm belt upgrade on my machine a while before before upgrading to XL.
Yes, I should have put that in my reply, that kevlar was your preference, sorry about that…
I have all 3 to try so I will start with the steel and see how that goes. I am still wary about the kevlar as my setup isn’t well controlled for humidity (or temp, but more so for humidity)…
1 Like
Griff
(Well crap, my hypometric precursor device is blown…)
49
Indeed.
I’ve tried ‘em all over the past 4 years.
In my experience steel have “failed” (distort, tell me to change them out) most quickly. Especially at “guitar string tight”. I’m not an engineer nor a machinist, I have no way to quantify, just my practical experience.
I’m progressing into year two on Kevlar belts, Not too tight, not too loose.
Now, if I could just keep the rails clean under the v-wheels!
Griff
(Well crap, my hypometric precursor device is blown…)
51
Yes and no. That brilliant piece of work was subsequent to my abandonment of steel belts. BUT, I stopped striving for “guitar string tight” belts upon breaking a stepper motor shaft a while back.
I find that a periodic check of the twang frequency quickly tells me if the belts have started to stretch and go slack. The only time I’ve seen this so far on the kevlar belts is after some abuse though…
3 Likes
Griff
(Well crap, my hypometric precursor device is blown…)
53
Same here. If the belts seem a bit stretched but appear un damaged I use this:
