Notifications
Clear all

Filament friction of MMU2s solution  

  RSS
dyits
(@dyits)
Active Member
Filament friction of MMU2s solution

My mmu2s fuction is good, but I alomast do not use it. Because the filament friction through the tube is big,especilly, between extuder and mmu.Then, filamant break often during printing or print quality is so bad. I find the motor gear of mmus is not running during pringting.

So is it possible to make the motor gear of mmu run synchronization with extruder motor during printing.  or is there a version of firmware that can do this.

Posted : 23/03/2020 2:53 am
kUMters
(@kumters)
Estimable Member
RE: Filament friction of MMU2s solution

Change that ptfe tubes to version with 3mm inner diameter 

MK3S + MMU2S (@BMG + MOSQUITO), MINI, ANET A8...
Posted : 23/03/2020 6:53 am
dyits
(@dyits)
Active Member
Topic starter answered:
RE: Filament friction of MMU2s solution

@kumters

Thank you, but I even try 4mm internal dimater. feels no change!

Posted : 24/03/2020 1:49 am
vintagepc
(@vintagepc)
RE: Filament friction of MMU2s solution
Posted by: @dyits

So is it possible to make the motor gear of mmu run synchronization with extruder motor during printing.  or is there a version of firmware that can do this.

Not with the current architecture, as I noted in the other thread where you asked this. And even if you started from scratch and rewrote the interface... any  delay between the two motions (which there will be, the MMU is a serial client) and any print with a bunch of close retractions is going to shred your filament in short order. You're constraining it to two points with such an implementation, there's nowhere in between for it to go or come from if there's any over or underfeed. 

Not to mention you'll probably bog down your Einsy since it would have to constantly communicate every E axis move to the MMU.

The sources are all on Prusa's github so you're welcome to try, but consider yourself warned it's going to be a Sisyphean task.

 

And, unless you changed all of the fittings there's no way you could have used 4mm ID PTFE. It would not have fit.

This post was modified 2 years ago by vintagepc
Posted : 24/03/2020 11:42 am
kUMters
(@kumters)
Estimable Member
RE: Filament friction of MMU2s solution

@dyits

OD=4mm and ID=4mm mean no PTFE/Bowden tube at all:) PS: if you use PTFE OD4/ID3 then there will be no friction in Bowden - it means your problem is somewhere else...

MK3S + MMU2S (@BMG + MOSQUITO), MINI, ANET A8...
Posted : 24/03/2020 11:46 am
djspathis
(@djspathis)
Trusted Member
RE: Filament friction of MMU2s solution

Does the exact length of the tube matter or might it be increased to help reduce the somewhat tight bending at extreme X Y print locations?

 

DJ...
Posted : 24/03/2020 2:25 pm
vintagepc
(@vintagepc)
RE: Filament friction of MMU2s solution
Posted by: @djspathis

Does the exact length of the tube matter or might it be increased to help reduce the somewhat tight bending at extreme X Y print locations?

 

It can be increased. The Mk3S uses the filament sensor to determine when it has loaded to the gears. You will see it stop at the normal distance and then feed smaller increments until done. You can calibrate the bowden length per the old method if desired to tell it the new default length and skip it feeding a bunch of small increments if your tube is significantly longer.

Posted : 24/03/2020 2:37 pm
pcweber
(@pcweber)
Estimable Member
RE: Filament friction of MMU2s solution

I cannot say that this is the solution, but in my opinion  your mmu should be able to push through the ptfe 3mm, mine is pushing thtough the ptfe with an ID of 2mm. I did; however, have a similar problem. What it turned out to be that the idler had minor warping so by not being straight the bearings met the gears with some space that shouldn't have been there. To check that the idler is indeed round is simple:

1. Open the mmu up

2. Turn the power off or disable steppers.

3. Spin by hand the idler and look at where the idler motor is attached.

If the idler body moves, then the problem may be that the idler has some sort of defect. Mine did, I printed another idler replacing the parts from one to another. The results were nothing less than spectacular. There is so much more torque on the input and output that the Bowden has no effect on the performance. A side benefit is that the filament ends are practically perfect, better than the manual. Probably this is due to the filament being ejected effectively by the mmu. Of course, I have dialed the temperatures in as well.

 

Phil

 

Stay safe and healthy, Phil...
Posted : 30/03/2020 9:11 pm
Share: