Parametric Auto-Rewind Spool Holder  

Re: Parametric Auto-Rewind Spool Holder

After some customization/modification to the scad files I have mine set up and they are working great! Some notes:

1) As you can see several rolls are at their bitter end yet the spring always engaged, no issues.
2) I removed the 'fingers' and just printed the hub at spool width using a second spring retainer and spacer in a grooved axle.
3) I added code to print the spool hub width/depth so I could tell them apart while printing drafts and just kept it for final prints.
4) I made holders/stands that have front/back mount grooves to minimize the width of entire assembly (It is 530 mm using 71mm hubs).
5) I used the PRUSA PTFE tube pivot piece and bearings to maximize reuse.

Re: Parametric Auto-Rewind Spool Holder

I could be completely messed up but after some experimentaion and thought, I see some issues:

1. Plastic can never make a lasting spring. It creeps proportional to stress, time and temperature. Fiberglass reinforced would work if you could make it. Clock springs are made for this purpose, ie spring steel.

2. When the MMU is unloading, in order to keep the fiber from unraveling on the reel for a given fiber acceleration the torque on the spring must be greater than the torque created by the angular acceleration of the mass of the accelerating fiber and reel. If the forces match the reel is just tight. If the force is to small a momentary but noticeable slack on the reel occurs. (enough to cause a typical Amazon reel tangle) The only way to reduce this is either to decrease the acceleration or increase the torque. However, when the MMU later loads the fiber it must now overcome not only the angular acceleration of the reel but also the torque on the spring which means if the fiber acceleration is the same as unloading the tension or force on the fiber in the gears will double.

This is not insurmountable , it just means that the optimum unloading acceleration will always be around half the loading acceleration. Further the unloading speed will be much slower than that of the current or mmu2s sandwich and one-way mechanism which is not acceleration limited in the same way. The current Prusa MMU2 or MMU2S should be able to load with twice the acceleration by the same argument since it doesn't have to overcome the spring.

3. The spring torque above is controlled by a slip clutch which I didn't investigate but can't be ignored. It needs to be adjustable and stable. I did my experiments with a constant torque DC motor and still found the above issues.

Re: Parametric Auto-Rewind Spool Holder

I use Vincent's parametric auto rewind spool holder for a while now and it is definitivly the best solution for my wall monted spool holder setup.
To reverse strain in the spring, time to time I reverse the spool assembly on the support, It works well anyway.
I do not use the middle hub that comes with his design, I made spacers inserted firmly into the reel.
To avoid tangling I sanded hub and spacer to obtain a smoother surface between them, since I didn't get any tangle.

Re: Parametric Auto-Rewind Spool Holder

Unfortunately I had to stop using this design because every time the filament would unload the filament would not rewind all the way and instead fall off the spool, regardless of my tension setting or any tweak I did.
Found it to be more trouble than its worth...

Re: Parametric Auto-Rewind Spool Holder

I use Vincent's parametric auto rewind spool holder for a while now and it is definitivly the best solution for my wall monted spool holder setup.
To reverse strain in the spring, time to time I reverse the spool assembly on the support, It works well anyway.
I do not use the middle hub that comes with his design, I made spacers inserted firmly into the reel.
To avoid tangling I sanded hub and spacer to obtain a smoother surface between them, since I didn't get any tangle.

nm, found it.