CoreBoxx Rewinder Tensioner Arm Mount Fix

CoreBoxx Rewinder Tensioner Arm Mount Rebuild

REVISION HISTORY:

10/19/25 - Initial Public Release:
Step File v 1.0
.3mf Model v 1.0
.3mf Prusa Slicer Project v 1.0

The What:

Scratch built rework of the Rewinder Tensioner Mount for CoreBoxx

The Why:
I had a several problems with the supplied tensioner mount piece for the rewinders.  

The supplied tensioner mount was only in the full bed STL, and the step version was from the original Filamentalist project the rewinders were derived from and didn't have threads for the PC4-M6's as they used a different sized fitting there.   So this required a from scratch rebuild to address.  

The Issues:
 

Hole Sizes

The holes for the mounting screws were too large, and would way too easily pull out when using the BluRolls supplied screws, as they were wood screw type and would tear out with barely any torquing at all as there wasn't enough material for those screws to bite into.

As the BluRolls kit came way short on the number of countersunk screws in the first place so I just ordered some real countersunk M3's to replace those wood screws for all the winders.   Even if I hadn't there just needed to be more material for these wood screws to work.   These are now sized for M3’s and will provide more material for the M3’s to bite into as you’re essentially using the M3’s as taps.    

Thread Sizes and Filament Hangup on threading

When the filament enters this part it very often gets “hung up”, either on the PC4-M6's face, or in the transition from the input feed.

This is due to two issues, first being the printed threads are just way too loose and tear out very easily when threading in the PC4-M6.   As a result the fitting sits very loosely in the part exacerbating the problem of trying to thread the filament.

The second is the transition between the threaded area to the constrained hex pathway is backwards, as this pathway was designed for a version of the Filamentalist rewinders where the part faces the opposite direction, so the filament comes in from the PC4-M6 end.
 

 The Changes: 

 1 - Resized the  chassis mounting holes 
 

This is to accommodate CS M3 bolts vs wood screw threads.   This gives substantially more material to “bite” into when using the M3 to cut threads in the part and allows for quite a bit more torque when tightening the chassis sides to the mount part.

2 - Resized the printed threads 

This is to provide more material for the PC4-M6 to bite into which allows it to seat fully (and tightly) to remove the facing surface filament hangup due to the fitting not being fully seated.

In fairness, Fusion (which the original was created in) treats all modeled threads as if they are to be machined, which kinna sux for 3D printing.   To address that you have to create custom thread profiles, and put them in the thread definitions subdirectory which then get wiped when you update Fusion and you have to reinstall them after every update, whigh is a huge PITA (although there is a plugin that helps with this somewhat).    The “printable” threads in this case are set to fully engage the the PC4 connectror’s M6 threads.   When installing the connector, make sure it’s perpendicular to the surface.  It’ll offer some resistance for the first couple turns, which will then decrease until you’ve hit the bottom of the travel where you’ll notice a marked increase in resistance.   Once you hit that point give it another 1/8 - ¼ turn and it will fully seat.

3 - Rework part filament path

 

The reason the filament is binding on threading is because the part is literally a mirror of the original Filamentalist design.   However in that design, the filament came in though the PC4-MX connector (path direction shown in cyan), which was fine.  Filament would come in, pass through the connector, go through the “purple box” corridor shown, pass through a reduction then follow a hex shaped constrained path till it exited the part.   

Because the way the Filmentalist reminders are being used in the CoreBoxx is the inverse of that, the path now follows the orange arrow, where it comes in from the right, passes through the “reducer” section (circled) and then has the entire length of the “purple box” corridor to deflect and plow into any of:

The PC4-M6's thread end face
The printed interface between the thread section and the “purple box” corridor
The gap between the PC4-M6 and the part if the connector is not fully seated due to the thread size issue resulting in the connector not being fully seated.

The original path portion looked like this:

Which left the entire area between the red and yellow circles as a wall for the filament to plow into:

To fix this I eliminated the reducer section,  and the entire “purple box” corridor and reset the end of thread interface to the rest of the part to be a flat surface so the PC4-M6 could fully seat and mate flat:

 

The filament infeed path from the right ends with the filament remaining constrained through the entire part right up until it comes within 0.2mm of the end of the PC4-M6, and because it's already centered it has no choice to go into the connector as the exit hole from the part is smaller by almost a factor of 2 to the entrance diameter of the threaded end of the PC4-M6 (again we're using this part ass backwards from it's normal usage).

One Possible Point of Binding:

All that remains in terms of potential issues is for the filament to get snagged on the leading edge of the PTFE going through the connector.   However if you add an internal chamfer this end of the PFTE, it will pretty much remove that as an issue as well.

FILES AND PRINTING NOTES:

When I post models to Printables I break the model up internally into a number of different objects. These are named in the step file appropriately and are in the (PS) Prusa Slicer variant of the .3mf file as well.  As the .3mf model file is a straight export from CAD the part names may not come through all the time.    Any needed supports via support enforcers are built into the model unless otherwise noted.

Part types are:

Model parts (which stay named based on the model)
Brims (which are also separate objects, and are going to be the same height as the first layer height (Typically 0.2mm), named as such in both (PS) .3mf's and Step Files.
Support Enforcers  THESE ARE NOT SUPPORTS. - See Below
Multi Color Parts will typically be named in a manner to make them easily identifiable for multi material prints. 

Part Revisions will be stamped where there is room to do so, and any revisions will be noted in the change log and in the component parts in the files.

ASA READY

I model all my parts as though they will be printed in ASA or other material that needs brims, even if I'm printing it in something else, in case later I might need to print in ASA.

As a result, the Brims are built in as a separate object.    If not printing in ASA just mark the brim object(s) as non printing and you're good to go.

SUPPORT ENFORCERS, NOT SUPPORTS 

I model any needed supports into the model but not as supports but support enforcers.

The reason for this is I always do supports as a multimateral print, and usually use PLA as the support interface material.   Note INTERFACE, not SUPPORT.   

The reason is PLA isn't super happy sitting directly on a 100-110C bed for a long print.   Therefore I use the main material for the supports a just use PLA (or whatever) as a support interface.   

If you aren't used to modeling this stuff in and typically just do paint on supports there is a difference.

Modeled in support enforcers HAVE to go THOUGH the surface of the model proper in order for PrusaSlicer to trigger the support generation under it.   That bit’s not very well documented.    
 

As a result the object is TALLER than the support itself will be (I always make them one “normal” layer height taller), so in this case it's 0.25mm taller than the support alone would be.

These enforcers are sized and placed to ensure the support structure is exactly where it needs to be and is sized cross sectionally to ensure it will NOT get hung up on the model.

Using The Enforcer as a Support

However if you aren't printing via multi material and just need it to act as a standard support that's easy enough to do:

Select the enforcer object in the hierarchy, uncheck the uniform scale lock to the left, and remove one normal layer height (0.25mm) from the Enforcer's Z size.   Once you do so the little “z grounding” widget icon will show up to the right of the Z position field.  Click that and it will drop the resized enforcer object to the deck, and due to removal of the overlapping size, you now have a standard support object, exactly sized and placed.