Ultimaker2 Feeder System - Improvements and Ideas

[alluvian 0]

You can move it to wherever you like just as long as the cable harness for the motor is long enough and you add spacers between the motor and extruder housing if you don't have the side panel in between.

As for friction, there's a fair amount of friction inside the actual extruder. I uploaded a few pictures a while ago but didn't make them public at first. Here they are:

 

Wow, after noticing that rubbing, did you do anything about that? Like file down the contact point or add a small bowden section to the entry end of the extruder to straighten out the entrance angle?

 

[IRobertI 520]

No, I haven't tried filing it down. I did try the bowden thing as... uhm, can't remember the name now, had good success with that. It was a long section though as I didn't want to cut up my only spare.

My progress since last time is that I've trimmed away some fat and shrunk things down a bit more. Still not ready for testing though as there's a few issues left.

 

I might flip it around and have the lever on the left side to avoid the whole sidepanel issue but I'm not sure how awkward it'll be to use then with the filament roll in the way. I'll have to print out a paper template and take a look I guess.

 

Oh, and I hope someone tells me if I'm doing something completely retarded...

 

[bertho 6]

How do you plan to latch it?

I like symmetrical designs since it keeps the forces symmetrical which helps especially since we are using low-strength materials.

Bertho

 

[IRobertI 520]

Still bouncing a few ideas around, haven't decided what I like yet. The problem as you say is the low strength of the material. An idea I like (that wont work) is a snap latch going from the top case out to the right. As you close the pressure assembly it pushes this arm to the side (towards you in the picture), it rides on the side of the arm and then snaps shut when the arm is in place. But, delamination...

I'm leaning towards something similar to the current design with a lever that goes down into a slit in the arm that sticks up (if you look closely I toyed with a cut-out for a latch there).

 

[gadgetfreak 7]

Hi guys, when re-designing can you please make sure to adapt the new design for flexible filament like discussed here:

http://umforum.ultimaker.com/index.php?/topic/4197-improving-the-um2-feeder-mechanism-for-flexible-filament/

I think that would be the easiest part of the new design.

It is really cool to print in flexible material!

See examples here - we printed a BMX handle:

http://www.youtube.com/watch?v=JIRrVSU0wnY&feature=youtu.be

and here - we helped architect print a mold for concrete:

https://scontent-a.xx.fbcdn.net/hphotos-prn1/t1/1017441_236844363165002_146607555_n.jpg'>https://scontent-a.xx.fbcdn.net/hphotos-prn1/t1/1017441_236844363165002_146607555_n.jpg'>https://scontent-a.xx.fbcdn.net/hphotos-prn1/t1/1017441_236844363165002_146607555_n.jpg'>https://scontent-a.xx.fbcdn.net/hphotos-prn1/t1/1017441_236844363165002_146607555_n.jpg

https://scontent-a.xx.fbcdn.net/hphotos-prn1/t1/1017441_236844363165002_146607555_n.jpg

 

Also Bertho, here is some info I think you requested over at Google Groups - regarding the direct extrude motor:

>>>Here's some related info from two weeks ago from Daid (numbers) and Joergen (sparkfun link):

>> It has a holding torque of 44 N*cm, 400 steps per revolution. NEMA17 of

>> 47mm. However, unless you are doing direct-drive it's way overpowered.

>> I have no idea where to get them in small numbers.

>google to the rescue: https://www.sparkfun.com/products/10846'>https://www.sparkfun.com/products/10846'>https://www.sparkfun.com/products/10846'>https://www.sparkfun.com/products/10846 (https://www.sparkfun.com/products/10846)

 

[IRobertI 520]

It has been in the back of my mind. I tried getting the "entrance" of the bowden as close to the idler as possible and I thought about adding a duplicate of the bowden mount on the entrance to the extruder so you could mount a small section of bowden there (although I doubt it's as important as the "entrance" since that's where it seems likely to get bent). But... for now, I want to try it my way first with minimum friction since I'm doing this for selfish purposes first and foremost

 

[bertho 6]

Great info! Thanks!

I have ordered the stepper and I also ordered the flexible material. Perfect timing since I need to do some rubber parts.

Thanks again,

Bertho

 

[IRobertI 520]

I think I'm ready for a first test print to find all the things that interfere and don't fit, what will break and so on. Would be nice to have a second extruder so I didn't have to pull the extruder apart every time I want to test something

 

[skint 61]

I really think there could be a better choice of feeder than a section of knurled bar. The rough knurled section is digging into and damaging the medium which cannot be a god thing! It seems sensible to change both the knurled bar and the opposite polished metal driving cylinder with rubber ones. Imagine two car wheels next to each other with the PLA filament held nicely in place by the two soft and flexible tyres. Without seeing it in action..I assume both the knurled bar and the opposite wheel both rotate but in opposite directions ?

Just a thought, wanted to let you know... Apologies for hijacking!

 

[ian 32]

I really think there could be a better choice of feeder than a section of knurled bar. The rough knurled section is digging into and damaging the medium which cannot be a god thing! It seems sensible to change both the knurled bar and the opposite polished metal driving cylinder with rubber ones. Imagine two car wheels next to each other with the PLA filament held nicely in place by the two soft and flexible tyres. Without seeing it in action..I assume both the knurled bar and the opposite wheel both rotate but in opposite directions ?

Just a thought, wanted to let you know... Apologies for hijacking!

 

not hijacking ! i think its a good idea and something i am quietly working on at the moment.

I am also not the biggest fan of using metal to tear up pla in the attempt to push it.. its just very brutal and not a clever idea.

I am thinking for in the direction of rubber from atleast two sides like you said. this way the system grabs the filament in the hand and gently pushes it up.. zero damage to the filament.... better ??

Ian :-)

 

[ian 32]

I think I'm ready for a first test print to find all the things that interfere and don't fit, what will break and so on. Would be nice to have a second extruder so I didn't have to pull the extruder apart every time I want to test something

 

Robert send Harma a quick PM or email (both!) tomorrow and ask for an extra extruder housing set.

It is a big help to have a working printer and a set to put your eyes on !!

Ian :-)

 

[IRobertI 520]

I have no idea if it would be significant but couldn't rubber introduce a slight error in extrusion? What I mean is that the rubber would give way ever so slightly under pressure so there would no longer be a true 1:1 relationship between the rotation of the motor and the movement of the filament. It's likely insignificant in the grand scheme of things but it was just something that popped into my head.

 

[ian 32]

your right about the 1 to 1 thing Robert.. true !!

but...if you had a good system where the rubber belt or wheel was under a movable spring system like your design ! then I think the slip factor would be relativly low... and .. this all has to be simply tested but I cant see at the moment a monster loss in quality from this...

PS we have to address the flow of filament into the extruder more... i can hear my pla bashing around the back of the ultimaker 2 right now and that aint good for smooth flow... :sad:

Ian :smile:

 

[bertho 6]

Robert Wrote:

The extrusion length will change with the rubber compression since the drive wheel radius will change. I do not know how significant it is and it could be compensated for.

Ian wrote :

I am thinking for in the direction of rubber from atleast two sides like you said. this way the system grabs the filament in the hand and gently pushes it up.. zero damage to the filament.... better ??

Ian :smile:

Gently?? it better grab it really tight :-)

 

[jhertzberg 19]

your right about the 1 to 1 thing Robert.. true !!

but...if you had a good system where the rubber belt or wheel was under a movable spring system like your design ! then I think the slip factor would be relativly low... and .. this all has to be simply tested but I cant see at the moment a monster loss in quality from this...

PS we have to address the flow of filament into the extruder more... i can hear my pla bashing around the back of the ultimaker 2 right now and that aint good for smooth flow... :sad:

Ian :smile:

 

It may be tough to fit, but how about using two belts in a miniature double conveyor, sort of like a double belt press, but symmetrical? The belts could be grooved down the middle to keep the filament on track.

 

[IRobertI 520]

Remember kids, leveling your bed is not optional... It's almost 3AM, I'm going to bed *sigh*

 

[alluvian 0]

Aw man! The parts in the background look so nice too, although not at the focal point. Did those two parts at least turn out okay?

 

[ian 32]

the two back parts look good Robert... ? so not a complete loss I guess :-)

 

[albert 2]

Very interesting thread!

What about a triangular design? looking at the chuck in my drill press...look at the way the drill is held in place by pressure from 3 sides at an angle of 120 degrees... if you could have pressure on the filament from 3 sides, the entrance channel would be a straight line, the filament could not deform, and you could open the extruder by removing or flipping open one of the 3 sides of it. Probably no need for a knurled surface on the drive wheels, just rough surfaces might provide enough friction. I realize that the dimensions would have to be very small. But the basic idea is to use as much of the available surface of the filament to apply pressure to from 3 sides.

However it is done I believe the best way to hold a round object is something with pressure from more than one side. Food for thought!

Albert

[albert 2]

hold a piece of filament in your fingers and watch how the soft tissue of the fingertips wraps around the entire surface of the filament.

3 hard rubber wheels ( think "pencil eraser") coming from 120 degree angles - their outer surfaces could touch so one wheel drives all three of them. they grab the filament completely. DECREASE pressure, INCREASE contact surface. If necessary repeat and make 6 wheels. What do you think?

[ian 32]

i like !!

can do you a quick pencil sketch of your idea ?

Ian

 

[ian 32]

this is the kind of idea.. i wonder what the best way would be to transfer energy to three sides ?? or only one side is powered and the two other sides are not ?

 

[drayson 75]

I think, to drive not only one, but two or three "movers" might be a good idea to decrease slip and increase contact surface.

Reading through all your posts and thinking of an direct drive bowden mod for my UM1 (as I don´t really like the space-needing solution for 2 extruders), I was considering two MK8 wheels next to each other guiding the filament left/right and a setup with a pre-loaded lever including a thumb release.

Just the distance of the two wheels and the drive is a bit difficult as it is dependend on filament diameter and play in the gears or chain/belt...

Might this be a possible way...?

 

[albert 2]

30 cent 120 degree extruder idea. See new album in photo gallery...how the heck can you get a photo in here?

[ian 32]

albert you have to click add media from gallery, then select something from your gallery.. in all honesty the forum system for attaching photos is mentally bad... i just add photos to a free blog site and then use the link anywhere anytime.. zero stress..