Notifications
Clear all

Belt Tension - possible solution?  

  RSS
chmax
(@chmax)
Eminent Member
Belt Tension - possible solution?

Hi,
I've been trying to fine tune my belts (being at 292/295) as, even if my print quality seems to be quite good, the recommended range is 240+/-40.

In thing forum I've seen pretty creative suggestions like tuning the belt as if it were a guitar string, the assembly manual gives some very subjective guidance, and I was wondering if a more scientific method couldn't be found. So here is my suggestion: could please all of you with good tension values measure( using a piece of string) the length of the belt from the wall of the extruder carriage to the other side (especially the guys from Prusa could deliver here some reliable values). This should be easy to do as it doesn't need any disassembly.

Please list your values like this:

model,AXIS (x/y), Tension Value for the axis (from the support menu), belt length (wall to wall in mm), hours of print time (in statistics menu)
ex: mk3,X,295,502,34

In this way, with a decent number of measurements, it should be possible to have a "golden length" that should produce "new" and "worn in" values (allowing for the belt to stretch a little after a while) and even put this in the assembly manual.

I'd really appreciate if you could help me trying this method....

Posted : 31/01/2019 11:47 am
Inflamedsebi
(@inflamedsebi)
Active Member
Re: Belt Tension - possible solution?

There are adjustable belt holder parts for y on thingyverse, where you can fine tune your belt by a screw.
For X there is already such a mechanic build in. Just have a look here.

just re-run the self test to make the printer update those values.

Posted : 31/01/2019 3:11 pm
thrawn86
(@thrawn86)
Honorable Member
Re: Belt Tension - possible solution?

its worth noting that the values it gives you are not intended to be any kind of scientific measurement, they're just a guideline calculated using some math voodoo from the stepper current response during calibration.

240 will likely be very different from 1 machine to another.

Posted : 31/01/2019 10:48 pm
Share: