From what i’ve been doing with:
Everything’s better under the C (a quick poke at the frame flex value)
and
What a difference 1/10th of a millimetre makes (how I got retraction to work at force 3/4/500)
I’ve done the retraction / extension dance a lot in a short period of time with different retraction force / machine state. I’ve observed:
- Technique gets it working, but mainly because it gives you leverage.
- After the belts have been retracted, you need to stretch the belt / find some slack in the system from somewhere.
- Fundamentally the retraction pre-tensions the belt with the amount of force it can exert with the
retraction force. - So you need to apply that much force + the amount of force needed to stretch the belt enough for the sensor to register.
- At low retraction forces, there can some slack in the way the belt is spooled that makes it easier.
- Fundamentally the retraction pre-tensions the belt with the amount of force it can exert with the
- A stiff / frictiony / seized spool makes it significantly worse:
- You have to overcome the spool friction + retraction force pre-tensioning (though the retraction force would have been a bit lower because of the friction)
- I guess its worse because the retraction from the motor has a mechanical advantage because of the gearing that’s higher than the mechanical advantage when rocking the attachment.
- Very speculative, but it also felt non-linear - extending from retracting at 600 required very little force, extending after retracting at 1800 felt like it needed loads, more than 3 times the force.